CN103642902B - Genetic analysis systems and method - Google Patents

Abstract

The invention provides the method being determined hereditary composite index score by the dependency assessed between individual genotype and at least one disease or state.This assessment comprises the Genome Atlas of individuality compared with confirming as the database of the medical science correlative heritability of being correlated with at least one disease or state.

Description

Genetic Analysis Systems and Methods

This application is a divisional application of the No. 200780050019.5 invention patent application with the filing date of November 30, 2007 and the invention title of "Genetic Analysis System and Method".

Background technique

Sequencing of the human genome and other recent advances in human genomics have revealed that the genome makeup between any two individuals is more than 99.9% similar. Relatively small amounts of variation in DNA between individuals are responsible for differences in phenotypic traits and are associated with many human diseases, susceptibility to various diseases, and response to disease treatments. Inter-individual variation in DNA occurs in both coding and non-coding regions, and includes base changes at specific sites in the genomic DNA sequence, as well as insertions and deletions of DNA. Variations that occur at a single base position in the genome are called single nucleotide polymorphisms, or "SNPs."

Although relatively rare in the human genome, SNPs account for the majority of the interindividual DNA sequence variation, occurring approximately every 1,200 base pairs in the human genome (see InternationalHapMapProject, www.hapmap.org). As more human genetic information becomes available, the complexity of SNPs is beginning to be understood. Consequently, the occurrence of SNPs in the genome is associated with the presence and/or susceptibility to various diseases and conditions.

As a result of these correlations and other advances in human genetics, medical and personal healthcare in general is moving toward a personalized approach in which a patient makes appropriate medical care taking into account his or her genomic information, among other factors choices and other choices. Accordingly, there is a need to provide individuals and their healthcare providers with information specific to that individual's personal genome to enable personalized medical and other decision-making.

Contents of the invention

The present invention provides a method for assessing the genotype correlation of an individual, the method comprising: a) obtaining a genetic sample of the individual, b) generating a genome map of the individual, c) combining the genome map of the individual with human genes determine the correlation of the individual's genotype to the phenotype by comparing it to a current database of correlations between the genotype and the phenotype, d) report the results obtained from step c) to the individual or to the individual's healthcare manager, e) when the When the additional human genotype correlation is known, the human genotype correlation database is updated with the additional human genotype correlation, f) by combining the genome map of the individual obtained by step c) or a part thereof with the additional human genotype The correlations are compared to update the individual's genotype correlations and determine additional genotype correlations for the individual, and g) report the results from step f) to the individual or to the individual's healthcare manager.

The present invention further provides a commercial method for assessing the genotype correlation of an individual, the method comprising: a) obtaining a genetic sample of the individual; b) generating a genomic profile of the individual, c) combining the genomic profile of the individual with comparison to a human genotype correlation database to determine the individual's genotype correlation; d) provide the individual with the results of determining the individual's genotype correlation in encrypted form; e) when additional human genotype correlations are known , updating the human genotype correlation database with the additional human genotype correlation; f) updating the individual's genotype correlation by comparing the individual's genome profile or a portion thereof with the additional human genotype correlation, and determining additional genotype correlations for the individual; and g) providing the individual or the individual's healthcare manager with the results of updating the genotype correlations for the individual.

Another aspect of the invention is a method of generating a phenotypic profile of an individual, the method comprising: a) providing a ruleset comprising rules, each rule indicating the relationship between at least one genotype and at least one phenotype b) providing a data set comprising genome profiles of each of a plurality of individuals, wherein each genome profile comprises a plurality of genotypes; c) regularly updating the rule set with at least one new rule, wherein the at least one new rule The rules indicate correlations between genotypes and phenotypes that were not previously associated with each other in the rule set; d) applying each new rule to the genome profile of at least one individual such that at least one genotype of the individual is associated with at least one associated with the individual phenotypes, and optionally, e) generating a report comprising the individual's phenotype profile.

The present invention also provides a system comprising: a) a rule set comprising rules, each rule indicating a correlation between at least one genotype and at least one phenotype; b) periodically using at least one new rule updating the code of the rule set, wherein the at least one new rule indicates a correlation between genotypes and phenotypes that were not previously associated with each other in the rule set; c) a database comprising a plurality of individual genome profiles; d) the rule collecting codes applied to an individual's genomic profile to determine an individual's phenotypic profile; and e) generating a report for each individual.

Another aspect of the present invention is that in the above-mentioned method and system, the transmission through the network is performed in an encrypted or unencrypted manner.

Introduced references

All publications and patent applications mentioned in this specification are herein incorporated by reference as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.

Specifically, the present invention relates to the following:

1. A method of assessing genotype correlation in an individual, the method comprising:

a) obtaining a genetic sample of said individual;

b) generating a genomic profile of said individual;

c) determining the genotype-phenotype correlation of the individual by comparing the genome profile of the individual to a current database of human genotype-phenotype correlations;

d) reporting said results obtained from step c) to said individual or said individual's healthcare manager;

e) updating said human genotype correlation database with additional human genotype correlations as they become known; and

f) updating said individual's genotype correlation by comparing said individual's genomic profile of step c) or a portion thereof with said additional human genotype correlation and determining said individual's additional genotype correlation ;and

g) reporting said results from step f) to said individual or said individual's healthcare manager.

2. The method of item 1, wherein a third party obtains the genetic sample.

3. The method of item 1, wherein said generating the genome map is performed by a third party.

4. The method of item 1, wherein the result is based on the GCI or GCIPlus score.

5. The method of clause 1, wherein said reporting comprises transmitting said results over a network.

6. The method of clause 1, wherein said reporting of said results is through an online portal.

7. The method of clause 1, wherein said reporting of said results is by paper or by e-mail.

8. The method of clause 1, wherein said reporting comprises reporting said results in an encrypted manner.

9. The method of clause 1, wherein said reporting comprises reporting said results in an unencrypted manner.

10. The method of item 1, wherein the genome profile of the individual is stored in an encrypted database or vault.

11. The method of clause 1, wherein the individual is a registered user.

12. The method of clause 1, wherein the individual is a non-registered user.

13. The method of item 1, wherein the genetic sample is DNA.

14. The method of item 1, wherein the genetic sample is RNA.

15. The method of item 1, wherein the genomic profile is a single nucleotide polymorphism genomic profile, the human genotype correlation database is a human single nucleotide polymorphism correlation, and the An additional human genotype correlation is a single nucleotide polymorphism correlation.

16. The method of item 1, wherein the genome profile includes truncations, insertions, deletions or duplications, the human genotype correlation database is human truncations, insertions, deletions or duplications correlations, and the Additional human genotype correlations are truncation, insertion, deletion or duplication correlations.

17. The method of item 1, wherein the genome profile is the entire genome of the individual.

18. The method of clause 1, wherein the method comprises evaluating 2 or more genotype correlations.

19. The method of clause 1, wherein the method comprises evaluating 10 or more genotype correlations.

20. The method of item 1, wherein the human genotype correlation database includes genetic variants in one or more genes listed in Table 1 and phenotypes associated with the genetic variants.

21. The method of item 1, wherein the human genotype correlation database includes genetic variants and genetic variants in one or more genes listed in Figures 4, 5, 6, 22 or 25 and the genetic variants associated with the genetic Variant-associated phenotypes.

22. The method of clause 1, wherein said human genotype correlation database includes genetic variants determined from said genomic profile of said individual and predetermined phenotypes exhibited by said individual.

23. The method of item 1, wherein the human genotype correlation database includes single nucleotide polymorphisms in the genes listed in Table 1 or Figures 4, 5, 6, 22 or 25 and phenotypes associated with said single nucleotide polymorphisms.

24. The method of clause 1, wherein the genetic sample is from a biological sample selected from the group consisting of blood, hair, skin, saliva, semen, urine, fecal material, sweat, and buccal samples.

25. The method of item 15, wherein the genotype correlation is the correlation of single nucleotide polymorphisms with diseases and conditions.

26. The method of item 15, wherein the genotype correlation is a correlation of a single nucleotide polymorphism with a phenotype of a non-medical condition.

27. The method of clause 1, wherein the genomic profile is generated using a high-density DNA microarray.

28. The method of clause 1, wherein the genomic profile is generated using genomic DNA sequencing.

29. The method of item 24, wherein the genetic sample is genomic DNA and the biological sample is saliva.

30. A method comprising:

a) providing a rule set comprising rules, each rule indicating a correlation between at least one genotype and at least one phenotype;

b) providing a data set comprising genomic profiles of each of the plurality of individuals, wherein each genomic profile comprises a plurality of genotypes;

c) periodically updating said rule set with at least one new rule, wherein said at least one new rule indicates correlations between genotypes and phenotypes that were previously uncorrelated with each other in said rule set; and

d) applying each new rule to said genomic profile of at least one of said individuals, thereby associating at least one genotype with at least one phenotype for said individual.

31. The method of item 30, further comprising:

e) generating a report comprising said phenotypic profile of said individual.

32. The method of item 30, further comprising: after step b)

i) applying said rules of said rule set to said genomic profile of said individual to determine a set of phenotypic profiles for said individual; and

ii) generating a report comprising the individual's initial phenotypic profile.

33. The method of clause 31 or 32, wherein providing the report comprises transmitting the report over a network.

34. The method of clause 31 or 32, wherein the report is provided in encrypted form.

35. The method of clause 31 or 32, wherein the report is provided unencrypted.

36. The method of clause 31 or 32, wherein the report is provided through an online portal.

37. The method of item 31 or 32, wherein the report is provided on paper or by email.

38. The method of item 30, wherein the new rule correlates uncorrelated genotypes with phenotypes.

39. The method of clause 30, wherein the new rules associate associated genotypes with phenotypes not previously associated with them in the rule set.

40. The method of clause 30, wherein the new rule changes a rule in the rule set.

41. The method of item 30, wherein said new rule is generated by correlation of a genotype from said genomic profile of said individual and a predetermined phenotype of said individual.

42. The method of item 30, wherein the rule associates multiple genotypes with a phenotype.

43. The method of clause 30, wherein applying the new rule further comprises determining the determined value based at least in part on a characteristic of the individual selected from the group consisting of race, ancestry, geography, sex, age, family history, and a predetermined phenotype. Describe the phenotype spectrum.

44. The method of item 30, wherein the genotype comprises nucleotide duplication, nucleotide insertion, nucleotide deletion, chromosomal translocation, chromosomal duplication, or copy number variation.

45. The method of item 44, wherein the copy number variation is a microsatellite duplication, a nucleotide duplication, a centromeric duplication, or a telomeric duplication.

46. The method of item 30, wherein the genotype comprises a single nucleotide polymorphism.

47. The method of item 30, wherein the genotype comprises a haplotype and a disomy.

48. The method of item 30, wherein the genotype comprises genetic markers of single nucleotide polymorphism linkage disequilibrium associated with the phenotype.

49. The method of item 30, wherein said phenotypic profile indicates the presence or risk of said quantitative trait.

50. The method of clause 30, wherein the phenotypic profile indicates the probability that an individual with a genotype has or will have a phenotype.

51. The method of clause 50, wherein the probability is based on a GCI or GCIPlus score.

52. The method of item 50, wherein the probability is an estimated lifetime risk.

53. The method of item 30, wherein said correlation is verified.

54. The method of clause 30, wherein the rule set includes at least 20 rules.

55. The method of clause 30, wherein the rule set includes at least 50 rules.

56. The method of clause 30, wherein said rule set includes rules based on said genotype correlations in Table 1.

57. The method of clause 30, wherein said set of rules includes rules based on said genotype correlations in Figures 4, 5, 6, 22, or 25.

58. The method of item 30, wherein the phenotype comprises a quantitative trait.

59. The method of item 58, wherein the quantitative trait comprises a medical condition.

60. The method of clause 59, wherein the phenotypic profile indicates the presence or absence of the medical condition, the risk of developing the medical condition, the prognosis of the medical condition, the therapeutic effect of the medical condition, or the response to treatment for the medical condition described.

61. The method of item 58, wherein the quantitative trait comprises a phenotype of a non-medical state.

62. The method of item 58, wherein the quantitative trait is selected from a physical trait, a physiological trait, a mental trait, an emotional trait, race, ancestry, or age.

63. The method of item 30, wherein the individual is a human.

64. The method of item 30, wherein the individual is non-human.

65. The method of item 30, wherein the individual is a registered user.

66. The method of clause 30, wherein the individual is a non-registered user.

67. The method of clause 30, wherein said genomic profile comprises at least 100,000 genotypes.

68. The method of clause 30, wherein the genomic profile includes at least 400,000 genotypes.

69. The method of item 30, wherein the genomic profile includes at least 900,000 genotypes.

70. The method of item 30, wherein the genomic profile comprises at least 1,000,000 genotypes.

71. The method of clause 30, wherein the genome map comprises a substantially complete whole genome sequence.

72. The method of clause 30, wherein the data set comprises a plurality of data points, wherein each data point relates to an individual and comprises a plurality of data elements, wherein the data elements comprise a unique identifier selected from the individual , genotype information, microarray SNP identifiers, SNPrs identifiers, chromosome positions, polymorphic nucleotides, quality metrics, raw data files, images, extracted intensity scores, physical data, medical data, race, pedigree, geography , gender, age, family history, known phenotype, demographic data, exposure data, lifestyle data, and behavioral data.

73. The method of clause 30, wherein the periodic updating and applying occurs at least once a year.

74. The method of clause 30, wherein providing the data set comprises obtaining a genomic profile of each of a plurality of individuals by:

i) performing a genetic analysis on a genetic sample obtained from said individual, and

ii) encoding the analysis in computer readable form.

75. The method of clause 30, wherein the phenotypic profile comprises a monogenic phenotype.

76. The method of item 30, wherein the phenotype comprises a polygenic phenotype.

77. The method of clause 30, wherein said report includes an initial phenotype profile.

78. The method of item 30, wherein the report includes an updated phenotype profile.

79. The method of clause 30, wherein said report further includes information about said phenotype of said phenotype profile selected from one or more of the following: preventive strategies, health information , therapy, symptom awareness, early detection programs, intervention programs and precise identification and sub-categorization of said phenotypes in said phenotype profile.

80. The method of item 30, further comprising:

e) adding a new genome profile of a new individual to the individual data set;

f) applying said rule set to said genomic profile of said new individual; and

g) generating an initial report of the phenotypic profile of said new individual.

81. The method of item 30, comprising:

e) adding a new genome profile of said individual;

f) applying said rule set to said new genomic profile of said individual; and

g) generating a new report of the individual's phenotype profile.

82. A system comprising:

a) a rule set comprising rules, each rule indicating a correlation between at least one genotype and at least one phenotype;

b) regularly updating the code of the rule set with at least one new rule indicating correlations between genotypes and phenotypes that were not previously correlated with each other in the rule set;

c) a database comprising the genome profiles of a plurality of individuals;

d) applying said rule set to said genomic profile of an individual to determine a code for a phenotypic profile of said individual; and

e) Code to generate reports for each individual.

83. The system of clause 82, wherein the report is transmitted over a network.

84. The system of clause 82, wherein the report is provided encrypted.

85. The system of clause 82, wherein the report is provided unencrypted.

86. The system of clause 82, wherein the report is provided through an online portal.

87. The system of clause 82, wherein the report is provided by paper or email.

88. The system of clause 82, further comprising code to notify said individual of a new or revised correlation.

89. The system of clause 82, further comprising code for notifying said individual of new or revised rules that can be applied to said genomic profile of said individual.

90. The system of clause 82, further comprising code for notifying said individual of new or revised prevention and health information regarding said phenotype of said phenotype profile of said individual.

91. A kit comprising:

a) at least one sample collection container;

b) instructions for use in obtaining samples from individuals;

c) instructions for accessing the genomic profile of said individual obtained from said sample through an online portal;

d) instructions for accessing the phenotypic profile of said individual obtained from said sample through an online portal; and

e) Packaging for delivering said sample collection container to said sample processing facility.

92. An online portal comprising a website enabling an individual to access said phenotype profile, wherein said website allows said individual to perform at least one of the following:

a) selecting said rules to apply to said individual's genomic profile;

b) View initial and updated reports on said website;

c) print initial and updated reports from said website;

d) saving initial and updated reports from the website to the individual's computer;

e) obtaining preventive and health information about the phenotypic profile of said individual;

f) access to online or telephone-connected genetic counseling;

g) extract information to share with physicians/genetic counselors; and/or

h) Acquisition of matching services and products offered.

93. The online portal of item 92, wherein the information is transmitted via a network.

94. The online portal of item 92, wherein the website is encrypted.

95. The online portal of item 92, wherein the website is unencrypted.

96. The online portal of clause 92, wherein said individual has one or more options pertaining to said level of confidentiality of the individual's information, or one or more portions thereof.

97. The online portal of item 92, wherein the phenotypic profile includes treatable medical states.

98. The online portal of item 92, wherein said phenotypic profile includes medical conditions for which there is no current preventive measure or current therapy.

99. The online portal of clause 92, wherein said phenotypic profile includes non-medical states.

100. A method of assessing an individual's risk of acquiring a condition, the method comprising:

a) obtaining the genotype of the individual;

b) determining the GCI or GCIPlus score from the genotype;

c) generate a report from the GCI or GCIPlus score; and

d) providing said report to said individual or said individual's healthcare manager.

101. A method of assessing an individual's risk of acquiring a state, the method comprising:

a) obtaining the genotype of the individual;

b) generating a genomic profile of said individual;

c) determining the individual's risk of acquiring status from said genome profile and genotype correlation database;

d) generate a report from c);

e) obtain new information from said individual;

f) determining new risks to the state of acquisition by introducing said new information;

g) generate a report from f); and

h) providing said report to said individual or said individual's healthcare manager.

102. A method of assessing an individual's risk of acquiring a state, the method comprising:

a) obtaining the genotype of the individual;

b) generating a genomic profile of said individual;

c) determining an individual's risk of acquiring a status from said genomic profile and genotype association database, wherein said risk is based on more than one SNP;

d) generate a report from c);

e) providing said report to said individual or said individual's healthcare manager.

103. The method of item 100, 101 or 102, wherein the genotype of the individual is obtained directly from the individual.

104. The method of item 100, 101 or 102, wherein the individual's genotype is obtained from a third party.

105. The method of clause 100, 101 or 102, wherein said providing is via network transmission.

106. The method of item 101, wherein said new information is obtained from a biological sample of said individual.

107. The method of item 101, wherein the new information is obtained from anthropometric measurements of the individual.

108. The method of item 101 or 102, wherein said risk is scored by GCI or GCIPlus.

109. The method of item 100 or 108, wherein the GCI or GCIPlus score includes a pedigree of the individual.

110. The method of item 100 or 108, wherein the GCI or GCIPlus score includes the sex of the individual.

111. The method of item 100 or 108, wherein the GCI or GCIPlus score includes factors specific to the individual, wherein the factors are not derived from the genotype.

112. The method of item 111, wherein the factors are selected from the group consisting of: the individual's place of birth, parents and/or grandparents, kinship lineage, location of residence, location of residence of ancestors, environmental conditions, known health conditions, Known drug interactions, household hygiene conditions, lifestyle conditions, diet, exercise habits, marital status, and body measurements.

113. The method of item 107 or 112, wherein the individual's physical measurements are selected from the group consisting of: blood pressure, heart rate, glucose level, metabolite level, ion level, weight, height, cholesterol level, vitamin level, blood count, Body mass index (BMI), protein levels and transcript levels.

114. A method of assessing an individual's risk of acquiring a state, the method comprising:

a) obtaining the genotype of the individual;

b) generating a genomic profile of said individual;

c) determining an individual's risk of acquiring Alzheimer's disease (AD), colorectal cancer (CRC), osteoarthritis (OA) or exfoliation glaucoma (XFG), wherein the risk for AD is based on rs4420638, for CRC is based on rs6983267, rs4911178 for OA and rs2165241 for XFG;

d) generate a report from c);

e) providing said report to said individual or said individual's healthcare manager.

115. The method of item 102, wherein the risk is determined by at least 3, 4, 5, 6, 7, 8, 9, 10, or 11 SNPs.

116. The method of item 102, wherein the risk is determined by at least 2 SNPs.

117. The method of item 116, wherein said risk is for obesity (BMIOB) and at least one of said at least 2 SNPs is rs9939609 or rs9291171.

118. The method of item 116, wherein said risk is for Graves' disease (GD) and at least one of said at least 2 SNPs is rs3087243, DRB1*0301DQA1*0501, or is associated with DRB1*0301DQA1* 0501's linkage disequilibrium.

119. The method of item 116, wherein said risk is for hemochromatosis (HEM) and at least one of said at least 2 SNPs is rs1800562 or rs129128.

120. The method of item 116, wherein the risk is for myocardial infarction (MI) and at least one of the at least 2 SNPs is rs1866389, rs1333049 or rs6922269.

121. The method of item 116, wherein said risk is for multiple sclerosis (MS) and at least one of said at least 2 SNPs is rs6897932, rs12722489, or DRB1*1501.

i22. The method of item 116, wherein said risk is for psoriasis (PS) and at least one of said at least 2 SNPs is rs6859018, rs11209026 or HLAC*0602.

123. The method of item 116, wherein said risk is for restless legs syndrome (RLS) and at least one of said at least 2 SNPs is rs6904723, rs2300478, rs1026732, or rs9296249.

124. The method of item 116, wherein the risk is for celiac disease (CelD) and at least one of the at least 2 SNPs is rs6840978, rs11571315, rs2187668, or DQA1*0301DQB1*0302.

125. The method of item 116, wherein the risk is for prostate cancer (PC) and at least one of the at least 2 SNPs is rs4242384, rs6983267, rs16901979, rs17765344, or rs4430796.

126. The method of item 116, wherein said risk is for lupus (SLE) and at least one of said at least 2 SNPs is rs12531711, rs10954213, rs2004640, DRB1*0301 or DRB1*1501.

127. The method of item 116, wherein the risk is for macular degeneration (AMD) and at least one of the at least 2 SNPs is rs10737680, rs10490924, rs541862, rs2230199, rs1061170, or rs9332739.

128. The method of item 116, wherein said risk is for rheumatoid arthritis (RA) and at least one of said at least 2 SNPs is rs6679677, rs11203367, rs6457617, DRB*0101, DRB1*0401 Or DRB1*0404.

129. The method of item 116, wherein the risk is for breast cancer (BC) and at least one of the at least 2 SNPs is rs3803662, rs2981582, rs4700485, rs3817198, rs17468277, rs6721996, or rs3803662.

130. The method of item 116, wherein said risk is for Crohn's disease (CD) and at least one of said at least 2 SNPs is rs2066845, rs5743293, rs10883365, rs17234657, rs10210302, rs9858542, rs11805303, rs1000113, rs17221417, rs2542151 or rs10761659.

131. The method of item 116, wherein said risk is for type 2 diabetes (T2D) and at least one of said at least 2 SNPs is rs13266634, rs4506565, rs10012946, rs7756992, rs10811661, rs12288738, rs8050136, rs1111875 , rs4402960, rs5215 or rs1801282.

Description of drawings

Figure 1 is a flow diagram illustrating the method aspects of the invention.

Figure 2 is an example of genomic DNA quality control measures.

Figure 3 is an example of hybridization quality control measures.

Figure 4 is a table of typical genotype correlations from published literature with tested SNPs and effect evaluations. A-I) Indicates the genotype correlation of a single locus; J) Indicates the genotype correlation of two loci; K) Indicates the genotype correlation of three loci; L) is the race and country abbreviation used in A-K Index; M) is the index of the abbreviated phenotype name (ShortPhenotypeName) in A-K, heritability and heritability reference.

Figures 5A-J are tables of typical genotype correlations with effect evaluations.

Figures 6A-F are tables of typical genotype correlations and estimated relative risks.

Figure 7 is a sample report.

Figure 8 is a diagram of a system for analyzing and transmitting genomic and phenotypic profiles over a network.

Figure 9 is a flow chart illustrating the business method aspects of the present invention.

Figure 10: Effect of prevalence assessment on relative risk assessment. Each curve corresponds to a different value of the allele frequency in the population, assuming a Hardy-Weinberg Equilibrium. The two black lines correspond to odds ratios of 9 and 6, the two red lines correspond to odds ratios of 6 and 4, and the two blue lines correspond to odds ratios of 3 and 2.

Figure 11: Effect of allele frequency assessment on relative risk assessment. Each curve corresponds to a different value of prevalence in the population. The two black lines correspond to odds ratios of 9 and 6, the two red lines correspond to odds ratios of 6 and 4, and the two blue lines correspond to odds ratios of 3 and 2.

Figure 12: Pairwise comparison of absolute values of different models.

Figure 13: Pairwise comparison of rank values (GCI scores) based on different models. The Spearman correlations between different pairs are given in Table 2.

Figure 14: Effect of popularity reporting on GCI scores. The Spearman correlation between any two popularity values is at least 0.99.

Figure 15: Diagram for an example web page from a personal portal.

Figure 16: Diagram of an example web page from a personal portal illustrating a personal's risk of developing prostate cancer.

Figure 17: Diagram of an example web page from a person portal illustrating a person's risk of developing Crohn's disease.

Figure 18: Histogram of HapMAP-based GCI scores for multiple sclerosis using 2 SNPs.

Figure 19: Individual lifetime risk of multiple sclerosis using GCIPlus.

Figure 20: Histogram of GCI scores for Crohn's disease.

Figure 21: Table for multi-locus correlations.

Figure 22: Table for SNP and phenotype correlations.

Figure 23: Table for phenotype and prevalence.

Figure 24: Glossary of abbreviations in Figures 21, 22 and 25.

Figure 25: Table for SNP and phenotype correlations.

detailed description

The present invention provides methods and systems for generating phenotype profiles based on stored genome profiles of individuals or groups of individuals, and conveniently generating original and updated phenotype profiles based on stored genome profiles. A genomic profile is generated by genotype determination from a biological sample obtained from an individual. A biological sample obtained from an individual can be any sample from which a genetic sample can be obtained. Samples can be from buccal swabs, saliva, blood, hair, or any other type of tissue sample. The genotype can then be determined from the biological sample. A genotype can be any genetic variant or biomarker, eg, single nucleotide polymorphisms (SNPs), haplotypes, or sequence of the genome. A genotype may be the entire genome sequence of an individual. Genotypes can be derived from high-throughput analysis generating thousands or millions of data points, eg, microarray analysis for most or all known SNPs. In other embodiments, genotypes can also be determined by high throughput sequencing.

The genotype forms the genome map of an individual. Genomic profiles are digitally stored and easily accessed at any point in time to generate phenotypic profiles. Phenotype profiles are generated by applying rules that correlate or combine genotypes with phenotypes. Rules can be developed based on scientific studies showing correlations between genotypes and phenotypes. Correlations may be curated or confirmed by a committee of one or more experts. By applying rules to an individual's genomic profile, associations between an individual's genotype and phenotype can be determined. An individual's phenotypic profile will have this certainty. The determination may be a positive correlation between an individual's genotype and a given phenotype, such that the individual has or will develop the given phenotype. Alternatively, it can be determined that an individual does not have or will not develop a given phenotype. In other embodiments, the determination may be a risk factor, estimate, or probability that the individual has or will develop the phenotype.

The determination can be based on a variety of rules, for example, a variety of rules can be applied to a genomic profile to determine the association of an individual's genotype with a particular phenotype. The determination process can also include individual-specific factors such as race, gender, lifestyle (eg, eating and exercise habits), age, environment (eg, location of residence), family medical history, personal medical history, and other known phenotypes. The incorporation of specific factors can be done by amending existing rules to include these factors. Alternatively, separate rules can be generated from these factors and applied to the phenotype determination of an individual after existing rules have been applied.

A phenotype can include any measurable trait or characteristic, such as susceptibility to a disease or response to drug treatment. Other phenotypes that may be included are physical and mental traits such as height, weight, hair color, eye color, sunburn sensitivity, size, memory, intelligence, optimism, general temperament. Phenotypes can also include genetic comparisons to other individuals or organisms. For example, an individual may be interested in the similarity between their genome profile and that of a famous person. They may also compare their genetic profile with other organisms such as bacteria, plants or other animals.

Taken together, the collection of related phenotypes determined for an individual constitutes that individual's phenotypic profile. Phenotype profiles can be accessed through an online portal. Alternatively, the phenotype profile may be provided in paper form as it existed at a particular time, with subsequent updates also provided in paper form. Phenotype profiles are also available through the online portal. The online portal may optionally be an encrypted online portal. Access to phenotype profiles may be provided to registered users who order rules for generating correlations between phenotypes and genotypes, determine an individual's genomic profile, apply rules to a genomic profile, and generate an individual's phenotype spectrum of individuals served. Access may also be provided to non-registered users, where they may have limited access to their phenotype profiles and/or reports, or may allow generation of initial reports or phenotype profiles, but updated reports only through a paid subscription . Healthcare managers and providers such as nurses, physicians, and genetic counselors may also have access to phenotypic profiles.

In another aspect of the invention, genomic profiles can be generated and digitally stored for both registered and non-registered users, but access to phenotypic profiles and reports can be limited to registered users. In another variation, both registered and non-registered users have access to their genotype and phenotype profiles, but non-registered users have restricted access or are allowed to generate limited reports, whereas registered users have full access and can Allows full report generation. In another embodiment, registered and non-registered users may initially have full access or complete initial reports, but only registered users may have access to updated reports based on their stored genomic profiles.

In another aspect of the invention, information on the association of multiple genetic markers with one or more diseases or conditions is combined and analyzed to obtain a genetic composite index (GCI) score. This score incorporates known risk factors as well as other information and assumptions, such as allele frequencies and disease prevalence. GCI can be used to quantitatively assess the association of a disease or state with the combined effect of a set of genetic markers. The GCI score can be used to provide a person untrained in genetics with a reliable (eg, robust), comprehensible, and/or intuitive view of their individual risk of disease compared to a relevant population based on existing scientific research. GCI scores can be used to generate GCIPlus scores. The GCIPlus score can include all GCI assumptions including risk (eg, lifetime risk), age-defined prevalence, and/or age-defined incidence by status. An individual's lifetime risk can then be calculated as a GCIPlus score proportional to the individual GCI score divided by the mean GCI score. The mean GCI score can be determined from a group of individuals with similar familial background, such as a group of Caucasians, Asians, East Indians, or other groups with a common familial background. The group may consist of at least 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55 or 60 individuals. In certain embodiments, the mean GCI score can be determined from at least 75, 80, 95 or 100 individuals. The GCIPlus score can be determined by determining an individual's GCI score, dividing that GCI score by the mean relative risk, and multiplying by the lifetime risk for the state or phenotype. For example, a GCIPlus score, such as in FIG. 19 , is calculated using data from FIG. 22 and/or FIG. 25 and information in FIG. 24 .

The invention includes the use of the GCI score described herein, and those skilled in the art will readily recognize the use of the GCIPlus score or variations thereof in place of the GCI score described herein.

In one embodiment, a GCI score is generated for each disease or condition of interest. These GCI scores can be aggregated to form an individual's risk profile. The GCI scores can be stored digitally so that they can be easily accessed at any point in time to generate a risk profile. The risk profile can be broken down by broad disease categories, eg, cancer, heart disease, metabolic disorder, mental disorder, bone disease, or ageon-set disorder. Large disease categories can be further broken down into subcategories. For example, for a large classification like cancer, one can e.g. by type (sarcoma, carcinoma or leukemia, etc.) or by tissue specificity (nerve, breast, ovary, testis, prostate, bone, lymph node, pancreas, esophagus, stomach, liver , brain, lung, kidney, etc.) lists the subcategories of cancer.

In another embodiment, an individual's GCI score is generated that provides readily understandable information about the individual's risk of acquiring or susceptibility to at least one disease or condition. In one embodiment, multiple GCI scores are generated for different diseases or states. In another embodiment, at least one GCI score can be accessed through an online portal. Alternatively, at least one GCI score may be provided in paper form, with subsequent updates also provided in paper form. In one embodiment, access to at least one GCI score is provided to a registered user who is an individual subscribing to the service. In an alternative embodiment, access is provided to non-registered users, where they may have limited access to at least one of their GCI scores, or they may allow at least one of their GCI scores to be generated Initial reports for items, but updated reports are only generated through paid subscriptions. In another embodiment, healthcare managers and providers, such as nursing staff, physicians, and genetic counselors, may also have access to at least one of an individual's GCI score.

There can also be a basic registration pattern here. Basic registration can provide phenotype profiles, where registered users can choose to apply all existing rules to their genomic profile, or a subset of existing rules to their genomic profile. For example, they may choose to only apply rules for actionable disease phenotypes. Basic registration can have different levels within the registration class. For example, different levels may depend on the number of phenotypes a registered user wants to associate with their genomic profile, or on the number of persons who have access to their phenotypic profile. Another level of basic registration could incorporate individual-specific factors, such as already known phenotypes such as age, sex, or medical history, into their phenotypic profile. Yet another level of basic registration may allow an individual to generate at least one GCI score for a disease or condition. A variant of this level may further allow an individual to specify the automatic generation of at least one GCI score for a disease or state if any change in the at least one GCI score is due to a change in the analysis used to generate the at least one GCI score. renew. In some embodiments, individuals may be notified of automatic updates by email, voice message, text message, postal mail, or facsimile.

Registered users can also generate reports with their phenotype profile and information about the phenotype, such as genetic and medical information about the phenotype. For example, the prevalence of the phenotype in the population, genetic variants for association, molecular mechanisms causing the phenotype, treatments for the phenotype, treatment options for the phenotype, and preventive actions may be included in the report. In other embodiments, the report may also include information such as similarity between the individual's genotype and the genotype of other individuals, such as celebrities or other well-known persons. Information about similarity can be, but is not limited to, percent homology, number of identical variants, and possibly similar phenotypes. These reports may further include at least one GCI score.

If the report is accessed online, the report may also provide links to other locations with further information about the phenotype, links to online support groups and message boards of people with the same phenotype or one or more similar phenotypes , a link to contact an online genetic counselor or physician, or a link to schedule a telephone or on-site appointment with a genetic counselor or physician. If the report is in paper form, the information could be the site location linked above or the phone number and address of the genetic counselor or doctor. Registered users can also choose which phenotypes to include in their phenotype profiles and which information to include in their reports. Phenotype profiles and reports can also be obtained by an individual's healthcare manager or provider, such as a nurse, physician, psychiatrist, psychologist, therapist, or genetic counselor. Registered users can also choose whether phenotype profiles and reports, or portions thereof, are obtained by the individual's healthcare administrator or provider.

The invention may also include a premium level of registration. Registered premium levels digitally maintain their genomic profiles after initial phenotypic profiles and reports are generated, and registered users are able to generate phenotype profiles and reports with updated correlations from recent studies. In another embodiment, registered users are able to generate risk profiles and reports with updated correlations from recent research. As research reveals new correlations between genotypes and phenotypes, diseases or states, new rules will be generated based on these new correlations and can be applied to already stored and maintained genomic profiles. New rules can correlate genotypes not previously associated with any phenotype, correlate genotypes with new phenotypes, revise existing correlations, or provide adjustments based on newly discovered associations between genotypes and diseases or conditions Basis for GCI Scoring. Registered users can be notified of new correlations by email or other electronic means, and they can choose to update their phenotype profile with the new correlations if the phenotype is of interest. Registered users can choose to pay per update, multiple updates within a specified time period (eg, 3 months, 6 months, or 1 year) or unlimited updates. Another level of registration could be that registered users have their phenotype profile or risk profile automatically updated whenever new rules are generated based on new correlations, rather than individuals choosing when to update their phenotype profile or risk profile. Risk profile.

On the other side of registration, registered users can introduce the following services to non-registered users: Generate correlation rules between phenotypes and genotypes, determine an individual's genomic profile, apply the rules to the genomic profile, and generate an individual's phenotype Spectrum. Registered users can refer registered users to preferential service customization prices or upgrade their existing registrations through introductions. Referral individuals may receive free access for a limited time or a discounted registration fee.

Phenotype profiles and reports and risk profile maps and reports can be generated for human and non-human individuals. For example, a subject can include other mammals such as cows, horses, sheep, dogs, or cats. As used herein, a Registered User is a human individual who subscribes to the Services by purchasing or paying for one or more Services. Services may include, but are not limited to, one or more of the following: determining a genomic profile of themselves or another individual (such as a registered user's child or pet); obtaining a phenotypic profile; updating a phenotypic profile and obtaining a genomic profile based on them and report on phenotype profiles.

In another aspect of the invention, a "field-deployed" mechanism can be derived from individual aggregations to generate individual phenotypic profiles. In preferred embodiments, an individual may have an initial phenotypic profile generated based on genetic information. For example, an initial phenotype profile including risk factors for different phenotypes and suggested treatment or preventive measures is generated. For example, a phenotype profile may include information on available drug treatments for a certain state and/or recommendations for dietary changes or exercise regimens. Individuals may choose to see or contact a physician or genetic counselor through a web portal or telephone to discuss their phenotypic profile. Individuals may decide to take a certain course of action, for example, take a particular drug treatment, change their diet, etc.

Individuals can then subsequently submit biological samples to assess changes in their physical state and possible changes in risk factors. Individuals can determine this change by submitting a biological sample directly to the agency that generated the genomic and phenotypic profiles (or a related agency, such as an agency contracted by the entity that generated the genetic profile and phenotypic profile). Alternatively, individuals may utilize a "regional deployment" mechanism in which individuals may submit their saliva, blood, or other biological samples to a testing device at their home, be analyzed by a third party, and the data transmitted for inclusion in another table. type spectrum. For example, an individual may receive an initial phenotypic report based on their genetic data reporting an increased lifetime risk of myocardial infarction (MI) to the individual. The report may also have recommendations for preventive measures to reduce the risk of MI, such as cholesterol-lowering drugs and dietary changes. Individuals may choose to contact a genetic counselor or physician to discuss the report and preventive measures and decide to change their diet. After a period of time on the new diet, individuals may visit their personal physician to have their cholesterol levels measured. New information (cholesterol levels) can be communicated (e.g., via the Internet) to entities with genomic information and used to generate new phenotypic profiles of individuals, as well as new risks for myocardial infarction and/or other conditions factor.

Individuals may also use "regional deployment" mechanisms or direct mechanisms to determine their individual response to specific drug treatments. For example, an individual can measure their response to a drug, and this information can be used to determine more effective treatment. Information that can be measured includes, but is not limited to, metabolite levels, glucose levels, ion levels (eg, calcium, sodium, potassium, iron), vitamins, blood counts, body mass index (BMI), protein levels, transcript levels, heart rate, etc. , this information can be determined by readily available methods and can be included in an algorithm to be combined with the initial genomic profile to determine a revised overall risk assessment score.

The term "biological sample" refers to any biological sample capable of being isolated from an individual, including a sample from which genetic material can be isolated. As used herein, "genetic sample" refers to DNA and/or RNA obtained from or derived from an individual.

As used herein, the term "genome" is used to refer to the entire set of chromosomal DNA found in the nucleus of a human cell. The term "genomic DNA" refers to one or more chromosomal DNA molecules, or a portion of a chromosomal DNA molecule, naturally present in the nucleus of a human cell.

The term "genomic profile" refers to a set of information about an individual's genes, such as the presence or absence of a particular SNP or mutation. A genomic profile includes an individual's genotype. A genomic profile may also be a substantially complete genome sequence of an individual. In some embodiments, the genomic map can be at least 60%, 80%, or 95% of an individual's complete genome sequence. A genome map can be approximately 100% of an individual's complete genome sequence. "A portion thereof" when referring to a genome map refers to a genome map of a subset of the genome map of the whole genome.

The term "genotype" refers to the specific genetic makeup of an individual's DNA. Genotypes can include genetic variants and genetic markers of an individual. Genetic markers and genetic variants may include nucleotide duplications, nucleotide insertions, nucleotide deletions, chromosomal translocations, chromosomal duplications, or copy number variations. Copy number variations may include microsatellite repeats, nucleotide repeats, centromeric repeats, or telomeric repeats. Genotypes can also be SNPs, haplotypes or diplotypes. Haplotypes can refer to loci or alleles. A haplotype can also be referred to as a statistically associated set of single nucleotide polymorphisms (SNPs) on a single chromatid. A diplotype is a group of haplotypes.

The term single nucleotide polymorphism or "SNP" refers to an expression of variation (eg, at least 1 percent point (1%)) on a chromosome relative to the identity of nitrogenous choline at a locus present in the human population. specific gene loci. For example, where one individual may have adenosine (A) at a particular nucleotide position in a given gene, another individual may have cytosine (C), guanine (G), or thymus at that position pyrimidine (T), so that the SNP is present at this specific position.

As used herein, the term "SNP genomic profile" refers to the base content of a given individual's DNA at SNP positions throughout the entire genome DNA sequence of the individual. "SNP profile" refers to the complete genome profile, or to a portion thereof, such as a more localized SNP profile that may be associated with a particular gene or a particular group of genes.

The term "phenotype" is used to describe a quantitative trait or characteristic of an individual. Phenotypes include, but are not limited to medical and non-medical states. Medical states include diseases and disorders. Phenotypes can also include physical traits such as hair color, physical traits such as lung capacity, mental traits such as memory retention, emotional traits such as anger control, ethnic traits such as ethnic background, familial traits such as an individual's place of birth, and traits such as Age profile of age expectation or age of onset for different phenotypes. Phenotypes can also be monogenic, where it is thought that one gene may be associated with the phenotype, or polygenic, where more than one gene is associated with the phenotype.

"Rules" are used to define correlations between genotypes and phenotypes. Rules can define correlations numerically, such as percentages, risk factors, or confidence scores. Rules can include multiple genotype-phenotype correlations. A "rule set" includes more than one rule. A "new rule" may be a rule indicating a correlation between a genotype and a phenotype for which a rule does not currently exist. New rules can link unlinked genotypes to phenotypes. New rules can also associate genotypes already associated with phenotypes with previously unassociated phenotypes. A "new rule" may also be an existing rule modified by other factors, including another rule. Existing rules can be modified due to known characteristics of the individual, such as race, ancestry, geography, gender, age, family history, or other previously determined phenotypes.

As used herein, "genotype correlation" refers to the statistical correlation between an individual's genotype (e.g., the presence of a mutation or mutations), and the predisposition to a phenotype (e.g., a particular disease, state , physical state and/or mental state). The frequency with which a particular phenotype is observed in the presence of a particular genotype determines the degree of genotype correlation or the likelihood that a particular phenotype will occur. For example, as detailed herein, a SNP leading to the E4 isoform of apolipoprotein is associated with the induction of early-onset Alzheimer's disease. A genotype correlation can also refer to a correlation or a negative correlation in which a phenotype is not predisposed. Genotype correlation can also represent an assessment that an individual has a phenotype or is predisposed to develop a phenotype. Genotype correlations can be represented by numerical values, such as percentages, relative risk factors, effect estimates, or confidence scores.

The term "phenotype profile" refers to the collection of phenotypes associated with a genotype or genotypes of an individual. A phenotypic profile may include information generated by applying one or more rules to a genomic profile or information about genotype correlations applied to a genomic profile. Phenotype profiles can be generated by applying multiple genotype-phenotype association rules. A probability or estimate can be expressed as a numerical value, such as a percentage, a numerical risk factor, or a numerical confidence interval. Probability can also be expressed as high, medium or low. Phenotype profiles can also indicate the presence or risk of a phenotype. For example, a phenotypic profile could indicate the presence of blue eyes or a high risk of developing diabetes. Phenotypic profiles can also indicate predicted prognosis, treatment efficacy, or response to treatment of a medical condition.

The term risk profile refers to the collection of GCI scores for more than one disease or condition. The GCI score is based on the analysis of the association between an individual's genotype and one or more diseases or conditions. Risk distribution plots can display GCI scores grouped by disease category. Further, the risk distribution map can display information on how to predict changes in GCI score with individual age or adjustment for multiple risk factors. For example, the GCI score for a particular disease may take into account the effects of dietary changes or preventive measures taken (smoking cessation, medication, bilateral radical mastectomy, hysterectomy). The GCI score can be displayed as a numerical measure, a graphical display, auditory feedback, or a combination of any of the foregoing.

As used herein, the term "online portal" refers to sources of information that are readily accessible to individuals through computers and Internet sites, telephones, or other means that allow similar access to information. The online portal can be an encrypted website. The website may provide links to other encrypted and non-encrypted websites, such as links to encrypted websites with individuals' phenotype profiles or links to non-encrypted websites such as message boards for individuals sharing a particular phenotype.

The practice of the present invention will employ, unless otherwise indicated, conventional techniques and protocols of molecular biology, cell biology, biochemistry and immunology, which are within the skill of the art. These conventional techniques include nucleic acid isolation, polymer array synthesis, hybridization, ligation, and hybridization detection using labels. Specific illustrations of suitable techniques are illustrated herein and references are given. However, other equivalent conventional methods may also be used. Additional routine techniques and instructions for use can be found in the following standard laboratory manuals and literature: For example, Genome Analysis: A Laboratory Manual Series (Vols. I-IV) (Genome Analysis: A Laboratory Manual Series (Vols. I-IV)), PCR Primers: Experimental Laboratory Manual (PCR Primer: A Laboratory Manual), Molecular Cloning: Laboratory Manual (Molecular Cloning: A Laboratory Manual) (all from Cold Spring Harbor Laboratory Press (Cold Spring Harbor Laboratory Press)), Stryer, L. (1995) Biochemistry (Fourth Edition) Freeman , New York, Gait, "Oligonucleotide Synthesis: A Practical Approach" 1984, IRL Press, London, Nelson and Cox (2000), Lehninger, Principles of Biochemistry, Third Edition, W.H. Freeman Pub., New York, N.Y.; and Berg et al. (2002) Biochemistry, Fifth Edition, W.H. Freeman Pub., New York, N.Y., all of which are hereby incorporated by reference in their entirety.

The methods of the invention involve analyzing the genome profile of an individual to provide molecular information about the phenotype to the individual. As detailed herein, an individual provides a genetic sample that generates a personal genome profile. An individual's genome profile is queried for data on genotype correlations by comparing the genome profile to a database of established and validated human genotype correlations. Databases of established and validated genotype correlations can be derived from the peer-reviewed literature and further developed by a committee of one or more experts in the field (e.g., geneticists, epidemiologists, or statisticians). Criticize and verify. In a preferred embodiment, rules are formulated based on validated genotype correlations and applied to an individual's genomic profile to generate a phenotypic profile. The results of the analysis of an individual's genomic profile (phenotypic profile) are provided to the individual or the individual's healthcare manager along with interpretation and supportive information, thereby giving the ability to make individualized choices about the individual's healthcare.

The method of the present invention is described in detail in Figure 1, where a genome profile of an individual is first generated. An individual's genome profile will include information about an individual's genes based on genetic variation and genetic markers. Genetic variation is the genotype, which makes up the genome map. These genetic variations or genetic markers include, but are not limited to, single nucleotide polymorphisms, single and/or polynucleotide repeats, single and/or polynucleotide deletions, microsatellite repeats (small Quantitative nucleotide repeats), dinucleotide repeats, trinucleotide repeats, sequence rearrangements (including translocations and duplications), copy number variations (deletions and gains at specific loci), etc. Other genetic variations include chromosomal duplications and translocations as well as centromeric and telomeric duplications.

Genotypes can also include haplotypes and disomytypes. In some embodiments, a genomic profile can have at least 100,000, 300,000, 500,000, or 1,000,000 genotypes. In some embodiments, a genomic profile can be substantially the entire genome sequence of an individual. In other embodiments, the genomic map is at least 60%, 80%, or 95% of the complete genome sequence of the individual. A genome map can be approximately 100% of an individual's complete genome sequence. Genetic samples containing target material include, but are not limited to, unamplified genomic DNA or RNA samples or amplified DNA (or cDNA). The target substance may be a specific region of genomic DNA containing a genetic marker of particular interest.

In step 102 of Figure 1, a genetic sample of an individual is isolated from a biological sample of the individual. These biological samples include, but are not limited to, blood, hair, skin, saliva, semen, urine, fecal material, sweat, buccal, and various body tissues. In some embodiments, a tissue sample can be taken directly from an individual, for example an oral sample can be obtained by an individual swabbing the inside of their cheek. Other samples such as saliva, semen, urine, fecal matter or sweat may also be provided by the individual himself. Other biological samples can be taken by a healthcare professional such as a phlebotomist, nurse or doctor. For example, a blood sample can be drawn from an individual by a nurse. A tissue biopsy can be performed by a healthcare professional, and the kit can also be used by the healthcare professional to efficiently obtain a sample. A small cylindrical skin sample may be removed or a needle may be used to remove a small tissue or fluid sample.

In some embodiments, an individual is provided with a kit having a sample collection container for the individual's biological sample. Kits may also provide instructions for individuals to directly collect their own samples, eg, how much hair, urine, sweat or saliva to provide. The kit may also include instructions for the individual requesting that the tissue sample be taken by a healthcare professional. The kit may include a site where samples may be collected by a third party, eg, the kit may be provided to a healthcare facility that subsequently collects samples from the individual. The kit may also provide return packaging for delivery of the sample to a sample processing facility where genetic material is isolated from the biological sample (step 104).

Genetic samples of DNA or RNA can be isolated from biological samples according to any of several known biochemical and molecular biological methods, see, e.g., Sambrook et al., Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, New York) (1989). There are also several commercially available kits and reagents for isolating DNA or RNA from biological samples, such as those available from DNA Genotek, GentraSystems, Qiagen, Ambion, and other suppliers. Oral sample kits are readily available commercially, such as the MasterAmp ^™ BuccalSwab DNA Extraction Kit from Epicentre Biotechnologies, as well as kits for DNA extraction from blood samples, such as the Extract-N-Amp ^™ from SigmaAldrich. DNA from other tissues can be obtained by digesting the tissue with protease and heat treatment, centrifuging the sample and extracting unwanted material with phenol-chloroform, leaving the DNA in the aqueous phase. The DNA can then be further isolated by ethanol precipitation.

In a preferred embodiment, genomic DNA is isolated from saliva. For example, using DNA self-collection kit technology available from DNAGenotek, individuals collect saliva samples for clinical processing. Samples can be conveniently stored and shipped at room temperature. After delivery of the sample to an appropriate laboratory for processing, DNA is isolated by heat denaturation and protease digestion of the sample (usually at 50°C for at least 1 hour using reagents provided by the collection kit supplier). The samples were then centrifuged and the supernatant was subjected to ethanol precipitation. The DNA pellet was suspended in a buffer suitable for subsequent analysis.

In another embodiment, RNA can be used as a genetic sample. In particular, genetic variation in expression can be identified from mRNA. The term "messenger RNA" or "mRNA" includes, but is not limited to, pre-mRNA transcripts, transcript processing intermediates, mature mRNA ready for translation and transcription of a gene or genes, or nucleic acids derived from mRNA transcripts. Transcript processing can include splicing, editing and degradation. As used herein, a nucleic acid derived from an mRNA transcript refers to the nucleic acid for which the mRNA transcript or a subsequence thereof ultimately serves as a template for its synthesis. Thus, cDNA reverse transcribed from mRNA, DNA amplified from cDNA, RNA transcribed from amplified DNA, etc. are all derived from mRNA transcripts. RNA can be isolated from any of several body tissues using methods known in the art, for example RNA isolation from unfractionated whole blood using the PAXgene ^™ Blood RNA System available from PreAnalytiX. Typically, mRNA will be used to reverse transcribe cDNA, which is then used or amplified for genetic variation analysis.

Genetic samples are usually amplified from cDNA reverse transcribed from DNA or RNA prior to genomic profiling. DNA can be amplified by a variety of methods, many of which use PCR. See, for example, PCR Technology: Principles and Applications for DNA Amplification (PCR Technology: Principles and Applications for DNA Amplification) (Ed.H.A. Erlich, Freeman Press, NY, N.Y., 1992); , Academic Press, San Diego, Calif., 1990); Mattila et al., Nucleic Acids Res.19, 4967 (1991); Eckert et al., PCR Methods and Applications (PCRMethodsandApplications) 1, 17 (1991); PCR (Eds.McPherson et al., IRL Press, Oxford); and US Patent Nos. 4,683,202, 4,683,195, 4,800,159, 4,965,188, and 5,333,675, each of which is hereby incorporated by reference in its entirety.

Other suitable amplification methods include ligase chain reaction (LCR) (for example, Wu and Wallace, Genomics, 4, 560 (1989), Landegren et al., Science, 241, 1077 (1988) and Barringer et al., Genes, 89: 117 (1990)), transcriptional amplification (Kwoh et al., Proc. Natl. Acad. Sci. USA86:1173-1177 (1989) and WO88/10315), autonomous sequence replication (Guatelli et al., Proc. Sci.USA, 87:1874-1878 (1990) and WO90/06995), selective amplification of target polynucleotide sequences (US Patent No. 6,410,276), consensus sequence primer polymerase chain reaction (CP-PCR) ( U.S. Patent No. 4,437,975), random primer polymerase chain reaction (AP-PCR) (U.S. Patent Nos. 5,413,909, 5,861,245), nucleic acid-based sequence amplification (NABSA), rolling circle amplification (RCA), Multiple displacement amplification (MDA) (US Pat. Nos. 6,124,120 and 6,323,009) and circle-to-circle amplification (C2CA) (Dahl et al., Proc. Natl. Acad. Sci 101:4548- 4553 (2004)). (See US Patent Nos. 5,409,818, 5,554,517 and 6,063,603, each of which is hereby incorporated by reference). Other amplification methods that may be used are described in US Patent Nos. 5,242,794, 5,494,810, 5,409,818, 4,988,617, 6,063,603, and 5,554,517, and US Patent Application Serial No. 09/854,317, each of which is incorporated herein by reference.

Generating the genome map of step 106 is accomplished using any of several methods. Several methods are known in the art to identify genetic variation, and these methods include, but are not limited to, DNA sequencing by any of several methods, PCR-based methods, fragment length polymorphism analysis (limited Fragment length polymorphism (RFLP), cleavage fragment length polymorphism (CFLP)), hybridization methods using allele-specific oligonucleotides as templates (e.g., TaqManPCR method, invader method, DNA chip method), a method using a primer extension reaction, a mass spectrometry method (MALDI-TOF/MS method), etc.

In one embodiment, high density DNA arrays are used for SNP identification and profile generation. These arrays are commercially available from Affymetrix and Illumina (see Affymetrix 500KAssayManual, Affymetrix, Santa Clara, CA (incorporated by reference); humanHap650Y genotyping beadchip, Illumina, San Diego, CA).

For example, Affymetrix GenomeWideHumanSNPArray 6.0 can be used to generate SNP distribution maps by genotyping over 900,000 SNPs. Alternatively, more than 500,000 SNPs analyzed with a complete genome sampling can be identified by using the Affymetrix GeneChipHumanMapping500KArraySet. In these analytical methods, a subset of the human genome is amplified by a single-primer amplification reaction using restriction enzyme-digested, adapter-ligated human genomic DNA. As shown in Figure 2, the concentration of ligated DNA can then be determined. The amplified DNA is then fragmented and the quality of the sample is determined before proceeding to step 106 . If the sample meets PCR and fragmentation criteria, the sample is denatured, labeled, and subsequently hybridized to a microarray consisting of small DNA probes at specific locations on the coated quartz surface. The amount of label hybridized to each probe is monitored as a function of the amplified DNA sequence, yielding sequence information and ultimately SNP genotyping.

AffymetrixGeneChip500KAssay was used according to the manufacturer's directions. Briefly, isolated genomic DNA was first digested with NspI or StyI restriction endonucleases. The digested DNA was then ligated with NspI or StyI linker oligonucleotides annealed to NspI or StyI restriction DNA, respectively. The ligated adapter-containing DNA was then amplified by PCR to generate amplified DNA fragments between about 200 and 1100 base pairs, as confirmed by gel electrophoresis. PCR products meeting the amplification criteria were purified and quantified for fragmentation. PCR products were fragmented with DNaseI to achieve optimal DNA chip hybridization. After fragmentation, DNA fragments should be less than 250 bp and average 180 bp as confirmed by gel electrophoresis. Samples meeting the fragmentation criteria are then labeled with a biotin compound using terminal deoxynucleotidyl transferase. The labeled fragments were then denatured and then hybridized into the GeneChip250K array. After hybridization, arrays were stained prior to scanning in a three-step process consisting of streptavidin phycoerythrin (SAPE) staining followed by biotinylated anti- Antibody amplification step for streptavidin antibody (goat), and final staining with streptavidin phycoerythrin (SAPE). After labeling, the array is covered with array holding buffer and then scanned with a scanner such as the Affymetrix GeneChipScanner3000.

After scanning with the AffymetrixGeneChipHumanMapping500KArraySet, data analysis was performed according to the manufacturer's instructions, as shown in Figure 3. Briefly, raw data were acquired using the GeneChip Operating Software (GCOS). Data can also be obtained by using the Affymetrix GeneChipCommandConsole ^™ . After obtaining the initial data, the GeneChip genotyping analysis software (GTYPE) was used for analysis. For the purposes of the present invention, samples with a GTYPE callrate of less than 80% were excluded. The samples were then analyzed using the BRLMM and/or SNiPer algorithms for testing. Samples with a call rate of less than 95% for BRLMM or less than 98% for SNiPer were excluded. Finally, correlation analysis was performed and samples with SNiPer quality index less than 0.45 and/or Hardy-Weinberg p-value less than 0.00001 were excluded.

As an alternative or in addition to DNA microarray analysis, genetic variations such as SNPs and mutations can be detected by DNA sequencing. DNA sequencing can also be used to sequence a substantial portion or the entire genome of an individual. In general, commonly used DNA sequencing is based on polyacrylamide gel fractionation to resolve strand end fragment populations (Sanger et al., Proc. Natl. Acad. Sci. USA 74:5463-5467 (1977)). Alternative methods have been developed and continue to be developed to improve the speed and ease of DNA sequencing. For example, high-throughput and single-molecule sequencing platforms are available from 454 LifeSciences (Branford, CT) (Margulies et al., Nature, (2005) 437:376-380 (2005)), Solexa (Hayward, CA), Helicos BioSciences (Cambridge, MA) (US Application No. 11/167046 filed June 23, 2005) and Li-Cor Biosciences (Lincoln, NE) (US Application No. 11/118031 filed April 29, 2005) are commercially available , or are being developed by them.

After the individual's genome profile is generated in step 106, the profile is digitally stored in step 108, which may be digitally stored in an encrypted manner. The genomic profile is encoded in a computer readable format for storage as part of a data set, and can be stored as a database, where the genomic profile can be "saved" and accessed again at a later time. A dataset includes multiple data points, where each data point relates to an individual. Each data point can have multiple data elements. A data element is a unique identifier used to identify an individual's genomic profile. It can also be a barcode. Another data element is genotype information, such as SNP or nucleotide sequence of an individual's genome. Data elements corresponding to genotype information may also be included in the data points. For example, if the genotype information includes SNPs identified by microarray analysis, other data elements may include microarray SNP identification numbers, SNPrs numbers, and polymorphic nucleotides. Other data elements may be chromosomal positions for genotype information, quality measures for data, raw data files, images of data, and extracted strength scores.

Individual-specific factors such as physical data, medical data, race, ancestry, geography, gender, age, family history, known phenotypes, demographic data, exposure data, lifestyle data, behavioral data, and other known Phenotypes can also be included as data elements. For example, these factors may include, but are not limited to, an individual's: place of birth, parents and/or grandparents, kinship lineage, location of residence, location of ancestors' residence, environmental conditions, known medical conditions, known drug interactions, family Health conditions, lifestyle conditions, diet, exercise habits, marital status, and body measurements (eg, weight, height, cholesterol levels, heart rate, blood pressure, glucose levels, and other measurements known in the art). The aforementioned factors of an individual's relatives or ancestors (eg, parents and grandparents) may also be introduced as data elements and used to determine the individual's risk for a phenotype or condition.

Specific factors can be obtained from questionnaires or from the individual's healthcare manager. Information from the "saved" graph can then be accessed and used as desired. For example, in an initial assessment of an individual's genotype relatedness, the entirety of the individual's information (typically SNPs or other genomic sequences on or taken from the entire genome) will be analyzed for genotype relatedness. In subsequent analyses, all or part of the information from the stored or banked genomic profile can be accessed as needed or as appropriate.

Comparison of Genome Maps and Genotype Correlation Databases

In step 110, genotype correlations are obtained from the scientific literature. The genotypic relevance of a genetic variation is determined from an analysis of populations of individuals that have been tested for the presence or absence of one or more phenotypic traits of interest and genotype profiles. The alleles of each genetic variation or polymorphism in the genotype profile are then tested to determine whether the presence of a particular allele is associated with the trait of interest. Correlation analyzes can be performed by standard statistical methods and statistically significant associations between genetic variants and phenotypic traits are documented. For example, it may be determined that the presence of allele A1 of polymorphism A is associated with heart disease. As a further example, it may be found that the combined presence of allele A1 of polymorphism A and allele B1 of polymorphism B is associated with an increased risk of cancer. The results of the analysis can be published in the peer-reviewed literature, confirmed by other research groups, and/or analyzed by a committee of experts (eg, geneticists, statisticians, epidemiologists, and physicians), and can also be validated .

Examples of correlations between genotypes and phenotypes on which the rules applied to genomic profiles are based are shown in Figures 4, 5 and 6. For example, in Figures 4A and B, each row corresponds to phenotype/locus/ethnicity, where Figures 4C to I include further information on the correlation of each of these rows. As an example, the "phenotype name abbreviation" of BC in FIG. 4A is the abbreviation of breast cancer as noted in the index of phenotype name abbreviations in FIG. 4M . In the row BC_4 (which is the class name of the locus), the gene LSP1 is associated with breast cancer. As shown in Figure 4C, the published or functional SNP identified for this association was rs3817198, while the published risk allele was C and the non-risk allele was T. Published SNPs and alleles were identified through publications (eg, basic publications in Figures 4E-G). In the example of LSP1 of Figure 4E, the underlying publication is Easton et al., Nature, 447:713-720 (2007). Figures 22 and 25 further list the dependencies. The correlations in Figures 22 and 25 can be used to calculate an individual's risk for a state or phenotype, eg, to calculate a GCI or GCIPlus score. GCI or GCIPlus scores can also incorporate information such as the popularity of a state, as in FIG. 23 .

Alternatively, correlations can be formed from stored genomic profiles. For example, individuals with stored genomic profiles may also have stored known phenotypic information. Analysis of stored genomic profiles and known phenotypes can form genotype correlations. As an example, 250 individuals with stored genome profiles also had stored information for a previous diagnosis of diabetes. Their genomic profiles were analyzed and compared to a control group of individuals without diabetes. Individuals previously diagnosed with diabetes are then determined to have a higher rate of the particular genetic variant compared to the control group so that a genotype association can be drawn between the specific genetic variant and diabetes.

In step 112, rules are formed based on the proven correlations between genetic variants and specific phenotypes. For example, rules can be generated based on the interrelated genotypes and phenotypes listed in Table 1. Correlation-based rules can incorporate other factors such as gender (eg, Figure 4) or race (Figures 4 and 5) to produce effect estimates as in Figures 4 and 5 . Other metrics produced by the rules can assess the relative risk increase as in FIG. 6 . Effect estimates and estimated relative risk increases can be derived from, or calculated from, published literature. Alternatively, rules can be based on correlations generated from stored genomic profiles and previously known phenotypes. In some implementations, the rules may be based on the correlations in FIGS. 22 and 25 .

In preferred embodiments, the genetic variant is a SNP. Although SNPs occur at single loci, individuals who carry a particular SNP allele at one locus can often be predicted to carry particular SNP alleles at other loci. The association of a SNP with an allele that predisposes an individual to a disease or condition arises through linkage disequilibrium, in which alleles at two or more loci are nonrandomly associated more or less frequently in a population than Frequency expected to result from random formation by recombination.

Other genetic markers or variants (eg, nucleotide repeats or insertions) may also be in linkage disequilibrium with genetic markers that have been shown to be associated with particular phenotypes. For example, nucleotide insertions are associated with phenotypes, and SNPs are in linkage disequilibrium with nucleotide insertions. Rules are formed based on the correlation between SNPs and phenotypes. Rules based on correlations between nucleotide insertions and phenotypes can also be formed. Either rule or both rules can be applied to genomic profiles, since the presence of one SNP can give a certain risk factor, the other can give another risk factor, and when combined they can increase risk.

By linkage disequilibrium, disease-prone alleles cosegregate with specific alleles of SNPs or combinations of specific alleles of SNPs. Specific combinations of SNP alleles along a chromosome are called haplotypes, and the regions of DNA where they combine can be called haplotype segments. Although a haplotype segment can consist of one SNP, a typical haplotype segment represents a series of 2 or more contiguous SNPs that exhibit low haplotype diversity among individuals and typically have low recombination frequencies. Identification of a haplotype can be performed by identifying one or more SNPs located in a haplotype block. In this way, often SNP profiles can be used to identify haplotype blocks rather than having to identify all SNPs in a given haplotype block.

Genotypic correlations between SNP haplotype patterns and diseases, conditions or physical states are increasingly known. For a given disease, the haplotype pattern of a group of people known to have the disease is compared to a group of people without the disease. By analyzing many individuals, the frequency of polymorphisms in a population can be determined, and these frequencies or genotypes can then be correlated with a particular phenotype (eg, disease or condition). Examples of known SNP-disease associations include polymorphisms of complement factor H in age-related macular degeneration (Klein et al., Science, 308:385-389, (2005)) and obesity-associated proximity to INSIG2 Variation of genes (Herbert et al., Science, 312:279-283 (2006)). Other known SNP associations include, for example, polymorphisms in the 9p21 region including CDKN2A and B (eg, rs10757274, rs2383206, rs13333040, rs2383207, and rs10116277 associated with myocardial infarction (Helgadottir et al., Science, 316:1491-1493( 2007); McPherson et al., Science, 316: 1488-1491 (2007)).

SNPs can be functional or non-functional. For example, a functional SNP has an effect on cellular function, resulting in a phenotype, whereas a non-functional SNP is functionally silent but can be in linkage disequilibrium with a functional SNP. SNPs can also be synonymous or non-synonymous. Synonymous SNPs are SNPs in which different forms result in the same polypeptide sequence, and are non-functional SNPs. SNPs are non-synonymous and can be functional or non-functional if they result in different polypeptides. SNPs or other genetic markers used to identify haplotypes in a diplotype (which is 2 or more haplotypes) can also be used to correlate phenotypes associated with the diplotype. Information about an individual's haplotypes, disomytypes, and SNP profiles can be in the individual's genomic profile.

In a preferred embodiment, for a rule based on genetic marker generation in linkage disequilibrium with another genetic marker associated with a phenotype, the genetic marker may have an ^r2 or D' score greater than 0.5, which is commonly used in the art used to determine linkage disequilibrium. In preferred embodiments, the score is greater than 0.6, 0.7, 0.8, 0.90, 0.95 or 0.99. Consequently, in the present invention, the genetic markers used to correlate a phenotype with an individual's genomic profile can be the same as or different from the functional or published SNPs associated with the phenotype. For example, using BC_4, the tested SNP and the published SNP are the same, just as the tested risk and non-risk alleles are the same as the published risk and non-risk alleles (Figure 4A and C). However, for BC_5, CASP8, and their association with breast cancer, the tested SNPs were different from their functional or published SNPs, just as the tested risk and non-risk alleles were to the published risk and non-risk alleles. Tested and published alleles are oriented relative to the positive strand of the genome, and from these columns a homozygous risk or non-risk genotype can be inferred, which can generate rules for, for example, a registered user's individual's genomic profile. In some embodiments, instead of identifying a test SNP, an allelic difference or SNP can be identified based on another analytical method (eg, TaqMan) using published SNP information. For example, for AMD_5 in Figure 25A, the published SNP is rs1061170, but no test SNP was identified. Test SNPs can be identified by LD analysis of published SNPs. Alternatively, instead of using a test SNP, TaqMan or other equivalent assays can be used to evaluate the genomes of individuals with the test SNP.

Test SNPs can be "DIRECT" or "TAG" SNPs (Figure 4E-G, Figure 5). A direct SNP is the same test SNP as a published or functional SNP, eg for BC_4. Direct SNPs can also be used for FGFR2 association in breast cancer, using SNPrs1073640 in Europeans and Asians, where the minor allele is A and the other allele is G (Easton et al., Nature, 447:1087-1093( 2007)). Another published or functional SNP of FGFR2 association also in breast cancer in Europeans and Asians is rs1219648 (Hunter et al., Nat. Genet. 39:870-874 (2007)). A tagging SNP is where the test SNP differs from a functional or published SNP, as in the case of BC_5. Tag SNPs can also be used for other genetic variants, for example, for CAMTA1 (rs4908449), 9p21 (rs10757274, rs2383206, rs13333040, rs2383207, rs10116277), COL1A1 (rs1800012), FVL (rs6025), HLA-DQA1 (rs4982888)5 , eNOS (rs1799983), MTHFR (rs1801133) and APC (rs28933380) SNPs.

Databases of SNPs are publicly available from, for example, the InternationalHapMapProject (see www.hapmap.org, The InternationalHapMapConsortium, Nature, 426.789-796 (2003), and The InternationalHapMapConsortium, Nature, 437:1299-1320 (2005)), Human Gene Mutation The Human Gene Mutation Database (HGMD) public database (see www.hgmd.org) and the Single Nucleotide Polymorphism database (dbSNP) (see www.ncbi.nlm.nih.gov/SNP/). These databases provide SNP haplotypes, or enable the determination of SNP haplotype patterns. These SNP databases thus enable the detection of genetic risk factors underlying a wide range of diseases and conditions such as cancer, inflammatory, cardiovascular, neurodegenerative and infectious diseases. These diseases or conditions may be treatable for which treatments and treatments currently exist. Treatment can include prophylactic treatment as well as treatment to ameliorate symptoms and condition, including lifestyle changes.

Many other phenotypes can also be detected, such as physical traits, physiological traits, mental traits, emotional traits, race, pedigree, and age. Physical traits may include height, hair color, eye color, body, or traits such as stamina, stamina, and agility. Mental traits may include intelligence, memory ability, or learning ability. Ethnicity and pedigree can include identification of pedigree or ethnicity, or where an individual's ancestry originated. Age may be a determination of the actual age of the individual, or the genetic characteristics of the individual that place it at that age relative to the general population. For example, an individual whose real age is 38 years old, but whose genetic characteristics determine his memory abilities or state of physical health may be an average of 28 years old. An additional age trait may be the life expectancy of the individual.

Other phenotypes may also include non-medical states, such as the "recreational" phenotype. These phenotypes can include comparisons to well-known individuals, eg, foreign aristocrats, politicians, celebrities, inventors, athletes, musicians, artists, business people, and individuals of notoriety (eg, criminals). Other "recreational" phenotypes may include comparisons to other organisms, eg, bacteria, insects, plants, or non-human animals. For example, an individual might be interested to see how their genome profile compares to that of their pet dog or a former president.

In step 114 , rules are applied to the stored genomic profile to generate a phenotype profile of step 116 . For example, the information in Figures 4, 5 or 6 may form the basis of rules or tests to be applied to an individual's genomic profile. The rules may include the information in Figure 4 about the tested SNPs and alleles and the effect estimates, where the UNITS of the effect estimates are the units of the effect estimates, eg OR, or odds ratio (95% confidence interval) or mean. In a preferred embodiment the effect assessment may be genotype risk (Fig. 4C-G), eg risk for homozygotes (homoz or RR), risk heterozygotes (heteroz or RN) and non-risk homozygotes (homoz or NN). In other embodiments, the effect estimate can be carrier risk, which is RR or RN versus NN. In yet other embodiments, effect assessment can be based on allelic, allelic risk, eg, R versus N. There is also genotype effect evaluation for two loci (Fig. 4J) or three loci (Fig. 4K) (eg, 9 possible genotype combinations for two locus effect evaluation: RRRR, RRNN, etc.). Test SNP frequencies in the public HapMap are also reported in Figures 4H and I.

In other embodiments, the information from Figures 21, 22, 23 and/or 25 can be used to generate information to apply to an individual's genomic profile. For example, the information can be used to generate a GCI or GCIPlus score for an individual (eg, Figure 19). Scoring can be used to generate information on the genetic risk (eg, estimated lifetime risk) of one or more states in an individual's phenotypic profile (eg, FIG. 15 ). The method allows calculation of estimated lifetime risk or relative risk for one or more phenotypes or states as listed in Figures 22 or 25 . Risk for a single condition can be based on one or more SNPs. For example, the estimated risk for a phenotype or state can be based on at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 SNPs, wherein the SNPs used to estimate risk can be published SNPs, Test for SNPs or both (eg, Figure 25).

The estimated risk for a state can be based on the SNPs listed in Figure 22 or 25. In some embodiments, the risk of a condition can be based on at least one SNP. For example, an individual's assessment of risk for Alzheimer's disease (AD), colorectal cancer (CRC), osteoarthritis (OA), or exfoliation glaucoma (XFG) can be based on 1 SNP (eg, rs4420638, rs6983267 for CRC, rs4911178 for OA and rs2165241 for XFG). For other states, such as obesity (BMIOB), Graves' disease (GD) or hemochromatosis (HEM), an individual's estimated risk can be based on at least 1 or 2 SNPs (e.g., rs9939609 and/or rs9291171 for BMIOB) ; DRB1*0301DQA1*0501 and/or rs3087243 for GD; rs1800562 and/or rs129128 for HEM). For conditions such as, but not limited to, myocardial infarction (MI), multiple sclerosis (MS), or psoriasis (PS), 1, 2, or 3 SNPs can be used to assess an individual's risk for these conditions (e.g., for MI is rs1866389, rs1333049, and/or rs6922269; for MS, rs6897932, rs12722489, and/or DRB1*1501; for PS, rs6859018, rs11209026, and/or HLAC*0602). To assess individual risk for restless legs syndrome (RLS) or celiac disease (CelD), 1, 2, 3, or 4 SNPs can be used (e.g., rs6904723, rs2300478, rs1026732, and/or rs9296249 for RLS; rs6840978, rs11571315, rs2187668 and/or DQA1*0301DQB1*0302). For prostate cancer (PC) or lupus (SLE), 1, 2, 3, 4, or 5 SNPs can be used to assess an individual's risk for PC or SLE (e.g., rs4242384, rs6983267, rs16901979, rs17765344, and/or rs4430796; for SLE rs12531711, rs10954213, rs2004640, DRB1*0301 and/or DRB1*1501). To assess an individual's lifetime risk of macular degeneration (AMD) or rheumatoid arthritis (RA), 1, 2, 3, 4, 5, or 6 SNPs (for example, rs10737680, rs10490924, rs541862, rs2230199, rs1061170 and/or rs9332739; for RA rs6679677, rs11203367, rs6457617, DRB*0101, DRB1*0401 and/or DRB1*0404). To assess an individual's lifetime risk of breast cancer (BC), 1, 2, 3, 4, 5, 6 or 7 SNPs (eg, rs3803662, rs2981582, rs4700485, rs3817198, rs17468277, rs6721996 and/or rs3803662) can be used. To assess an individual's lifetime risk of Crohn's disease (CD) or type 2 diabetes (T2D), 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or 11 SNPs (eg,对于CD是rs2066845、rs5743293、rs10883365、rs17234657、rs10210302、rs9858542、rs11805303、rs1000113、rs17221417、rs2542151和／或rs10761659；对于T2D是rs13266634、rs4506565、rs10012946、rs7756992、rs10811661、rs12288738、rs8050136、rs1111875、rs4402960、rs5215和/or rs1801282). In some embodiments, the SNPs used as the basis for the risk determination may be in linkage disequilibrium with the SNPs described above or listed in Figures 22 or 25 .

An individual's phenotypic profile can include many phenotypes. In particular, whether in symptomatic, pre-symptomatic or asymptomatic individuals (including carriers of one or more disease/state susceptibility alleles), patients are assessed for disease or other conditions ( For example, risk of possible drug response, including metabolism, efficacy and/or safety, enables prognostic or diagnostic analysis of susceptibility to a variety of unrelated diseases and conditions. Thus, these methods provide an overall assessment of an individual's susceptibility to a disease or condition without the need to preconceive any particular disease or condition test. For example, the methods of the invention enable the assessment of an individual's susceptibility to any of the various states listed in Table 1, Figures 4, 5 or 6 based on the individual's genomic profile. Furthermore, these methods allow for the estimation of lifetime risk or relative risk for individuals assessed for one or more phenotypes or states, such as those phenotypes in Figures 22 or 25 .

The evaluation preferably provides information on 2 or more of these states, and more preferably 3, 4, 5, 10, 20, 50, 100 or even more of these states. In a preferred embodiment, the phenotypic profile is obtained by applying at least 20 rules to the individual's genomic profile. In other embodiments, at least 50 rules are applied to the individual's genomic profile. A single rule of phenotype can be applied to a single gene phenotype. More than one rule may also be used for a single phenotype, such as a polygenic phenotype or a monogenic phenotype in which multiple genetic variants in a single gene affect the probability of that phenotype occurring.

After an initial scan of an individual patient's genomic profile, updating of individual genotype associations is performed (or employed) by comparison to additional nucleotide variants (eg, SNPs) as they become known. For example, step 110 may be performed on a regular basis, eg, daily, weekly or monthly, by one or more persons of ordinary skill in the field of genetics who search the scientific literature for correlations with new genotypes. The new genotype correlations can then be further confirmed by a committee of one or more experts in the field. Thereafter, step 112 may be periodically updated with new rules based on newly validated correlations.

New rules can include genotypes or phenotypes that are outside of existing rules. For example, a genotype not associated with any phenotype is found to be associated with a new or existing phenotype. The new rules can also be used for correlations between previously unavailable genotypes and their associated phenotypes. New rules can also be determined for genotypes and phenotypes that already have existing rules. For example, there are existing rules based on the correlation between genotype A and phenotype A. New research reveals that genotype B correlates with phenotype A, leading to new rules based on this correlation. Another example is finding that phenotype B correlates with genotype A, and making new rules accordingly.

Rules could be developed when discovering correlations based on known but not initially identified in the published scientific literature. For example, one might report that genotype C is associated with phenotype C. Additional publications report that genotype D is associated with phenotype D. Phenotypes C and D are related symptoms, eg phenotype C may be shortness of breath while phenotype D is smaller lung volumes. Genotype C and phenotype D or genotype D and phenotype can be discovered and confirmed by statistical methods using existing stored genome profiles of individuals with genotypes C and D and phenotypes C and D, or by further research Correlation between C. New rules can then be generated based on the newly discovered and confirmed correlations. In another embodiment, stored genotype profiles of a plurality of individuals with a particular or related phenotype can be studied to determine genotypes shared by these individuals, and to determine correlations. New rules can be generated based on this correlation.

Rules can also be made to amend existing rules. For example, a correlation between a genotype and a phenotype may be determined in part by known characteristics of the individual, such as race, ancestry, geography, sex, age, family history, or any other known phenotype of the individual. Rules based on these known individual characteristics can be formulated and incorporated into existing rules to provide revised rules. The choice of rule to apply the modification will depend on individual factors specific to the individual. For example, a rule might be based on a 35% probability that an individual has phenotype E when the individual has genotype E. However, if the individual is of a particular race, the probability is 5%. New rules can be formulated based on this result and applied to individuals with that particular racial identity. Alternatively, one could apply an existing rule that determines a value of 35%, and then apply another rule based on the racial characteristics of that phenotype. Rules based on known individual characteristics may be determined from the scientific literature or based on studies of stored genome profiles. When new rules are generated, they can be added in step 114 and applied to the genome map, or they can be applied periodically, for example at least once a year.

Information on individual risk of disease can also be expanded with technological advances in higher-resolution SNP genomic maps. As mentioned above, initial SNP genome distribution maps can be easily generated using microarray technology for scanning 500,000 SNPs. Assuming the case of haplotype segments, this number can be used for a typical distribution plot of all SNPs in an individual's genome. Nonetheless, it is estimated that about 10 million SNPs typically occur in the human genome (the International HapMap Project; www.hapmap.org). With technological advances in whole-genome sequencing or whole-genome sequencing that enable practical and economical resolution of SNPs at higher levels of detail (e.g., microarrays of 1,000,000, 1,500,000, 2,000,000, 3,000,000, or more SNPs), more detailed SNP genomes can be generated Distribution. Likewise, technological advances in computer analysis methods will enable the economic analysis of finer SNP genomic profiles and the updating of master databases of SNP-disease associations.

After the phenotype profile is generated at step 116 , the registered user or their healthcare manager can access their genomic or phenotype profile through an online portal or website as in step 118 . A report including the phenotype profile and other information about the phenotype profile and the genome profile may also be provided to the registered user or their healthcare manager, as described in steps 120 and 122 . Reports can be printed, stored on the registered user's computer, or viewed online.

Figure 7 shows an example online report. Registered users can choose to display a single phenotype or more than one phenotype. Registered users may also have different viewing options, for example, a "QuickView" option as shown in FIG. 7 . The phenotype can be a medical state and the different treatments and symptoms in the quick report can be linked to other web pages containing further information about the treatment. For example, clicking on a drug leads to a website that includes information on dosage, cost, side effects, and efficacy. Drugs can also be compared with other treatments. The website may also include a link to the website of the drug manufacturer. Another link could provide registered users with the option to generate a pharmacogenomic profile, which would include information on their likely response to drugs based on their genomic profile. Links to medication alternatives, such as preventive behaviors (such as fitness and weight loss); and links to dietary supplements, diet plans, and information on nearby health clubs, health clinics, health and links to rehabilitation providers, urban spas (dayspas), and more. Educational and informative videos, summaries of available treatments, possible therapies, and general advice may also be provided.

The online report may also provide a link to schedule a personal physician or genetic counseling appointment or to access an online genetic counselor or physician, thereby providing registered users with the opportunity to ask for more information about their phenotypic profile. Links to online genetic counseling and physician inquiries are also available on the online report.

Reports can also be viewed in other formats, such as aggregated observations for a single phenotype, which provide more detail for each category. For example, there may be more detailed statistics about the likelihood of a registered user developing a phenotype; more information about typical symptoms or phenotypes, such as symptoms representative of medical conditions or ranges for non-medical conditions of the body (such as height); or information about More information on genes and genetic variants, such as population prevalence, eg, in the world or in different countries, or in different age ranges or sexes. For example, Figure 15 shows a summary of estimated lifetime risks for a number of states. Individuals can view more information for a particular condition, such as prostate cancer (FIG. 16) or Crohn's disease (FIG. 17).

In another embodiment, the report may be a report of a "recreational" phenotype, eg, the similarity of an individual's genome profile to that of a well-known individual such as Albert Einstein. The report can show the percent similarity between the individual's genome profile and Einstein's individual's genome profile, and can further show Einstein's predicted IQ and the individual's predicted IQ. Further information may include how the genome profile and IQ of the overall population compares to the genome profile and IQ of the individual and Einstein.

In another embodiment, the report may display all phenotypes that have been associated with the registered user's genomic profile. In other embodiments, the report may only display phenotypes determined to be positively associated with the individual's genomic profile. Individuals may choose to display particular subclasses of phenotypes in other forms, such as only medical phenotypes or only treatable medical phenotypes. For example, addressable phenotypes and their associated genotypes may include Crohn's disease (associated with IL23R and CARD15), type 1 diabetes (associated with HLA-DR/DQ), lupus (associated with HLA-DRB1), Psoriasis (HLA-C), Multiple Sclerosis (HLA-DQA1), Graves' Disease (HLA-DRB1), Rheumatoid Arthritis (HLA-DRB1), Type 2 Diabetes (TCF7L2), Breast Cancer (BRCA2) , colon cancer (APC), episodic memory (KIBRA), and osteoporosis (COL1A1). Individuals can also choose to display subcategories of the phenotype in the report, for example, only inflammatory diseases for medical states or only physical traits for non-medical states. In some embodiments, an individual can select by highlighting those states for which estimated risk is calculated (e.g., Figure 15A, D), only states with higher risk (Figure 15B), or only states with lower risk (Figure 15C). States to display all the states for which an estimated risk has been calculated for that individual.

Information delivered and transmitted to individuals may be encrypted and confidential, and access to such information by individuals may be controlled. Information derived from complex genomic profiling can be provided to individuals as regulatory-approved, understandable, medically relevant, and/or high-impact data. Information may also be of general importance and not related to medical treatment. Encrypted transfer of information to an individual may be through several means including, but not limited to, portal interface and/or post. More preferably, the information is provided to the individual encrypted (if the individual so chooses) through a portal interface to which the individual has secure and confidential access. This interface is preferably provided through an on-line, Internet website portal, or alternatively, by telephone or other means that allow for private, secure and easy-to-use access. Genomic profiles, phenotypic profiles, and reports are provided to individuals or their healthcare managers through data transmission over the network.

Accordingly, FIG. 8 is a block diagram illustrating a representative example logic device by which phenotype profiles and reports may be generated. Figure 8 shows a computer system (or digital device) 800 for receiving and storing genomic profiles, analyzing genotype correlations, generating rules based on genotype correlations, applying rules to genomic profiles, and generating phenotype profiles and reports. The computer system 800 can be understood as a logical device capable of reading instructions from a medium 811 and/or a network port 805 which can optionally be connected to a server 809 with a fixed medium 812 . The system shown in FIG. 8 includes a CPU 801 , disk drive 803 , optional input devices (eg, keyboard 815 and/or mouse 816 ), and optional monitor 807 . Data communication with server 809 at a local or remote location can be accomplished through the communication media shown. Communication media may include any means of transmitting and/or receiving data. For example, a communication medium can be a network connection, a wireless connection, or an Internet connection. This connection can provide communication on the World Wide Web. It is envisioned that data pertaining to the present invention may be transmitted by these means over a network or connection for receipt and/or verification by party 822 . The recipient 822 may be, but is not limited to, an individual, a registered user, a healthcare provider, or a healthcare manager. In one embodiment, the computer-readable medium comprises a medium suitable for communicating the results of analysis of biological samples or genotype correlations. The medium may include results regarding a phenotypic profile of an individual subject, where such results are obtained using the methods described herein.

The Personal Portal will preferably be used as the primary interface for individuals receiving and evaluating genomic data. Portals will enable individuals to follow their samples from collection to testing and to track results. Accessed through the portal, individuals are informed about their relative risk of common genetic diseases based on their genomic profile. Registered users can choose which rules to apply to their genome map through the portal.

In one embodiment, one or more web pages will have a list of phenotypes and a box next to each phenotype that a registered user can select to include it in their phenotype profile. Phenotypes can be linked to information related to that phenotype to help registered users make informed choices about which phenotypes they wish to include in their phenotype profiles. Web pages may also have phenotypes organized by disease groupings (eg, treatable or non-treatable). For example, registered users can select only treatable phenotypes, such as HLA-DQA1 and celiac disease. Registered users can also choose to display phenotypes with presymptomatic or postsymptomatic treatment. For example, an individual may choose to have a manageable phenotype (in addition to further screening) with presymptomatic treatment, for celiac disease a presymptomatic treatment of a gluten-free diet. Another example could be Alzheimer's disease, presymptomatic treatments being statins, exercise, vitamins and psychoactive effects. Thrombosis is another example, and presymptomatic treatment is avoidance of oral contraceptives and avoiding prolonged sitting. An example of a phenotype with approved post-symptomatic treatment is wet AMD associated with CFH, where an individual can undergo laser therapy for their condition.

Phenotypes can also be organized by type or category of disease or condition, eg, neurological, cardiovascular, endocrine, immunological, etc. Phenotypes can also be grouped into medical and non-medical phenotypes. Other classifications of phenotypes on web pages can be by physical traits, physiological traits, mental traits or emotional traits. The webpage may further provide for selecting a partition of a set of phenotypes by selecting a box. For example, select all phenotypes, only medically relevant phenotypes, only non-medically relevant phenotypes, only treatable phenotypes, only non-treatable phenotypes, different disease groups, or "recreational" phenotypes. "Entertainment" phenotypes may include comparisons to celebrities or other well-known individuals, or comparisons to other animals or even other organisms. A list of genomic profiles available for comparison may also be provided on the web page for selection by the registered user to compare with the registered user's genomic profile.

The online portal may also provide a search engine to help registered users browse the portal, retrieve a specific phenotype, or retrieve specific terms or information revealed by their phenotype profiles or reports. Links to access companion services and offered products may also be provided by the portal. Additional links to support groups, message boards, and chat rooms for individuals with a common or similar phenotype may also be provided. The online portal may also provide links to other sites with more information about the phenotypes in the registered user's phenotype profile. The online portal may also provide services that allow registered users to share their phenotype profiles and reports with friends, family or healthcare administrators. Registered users can choose to display the phenotypes they wish to share with their friends, family or healthcare managers in the phenotype profile.

Phenotype profiles and reports provide individualized genotype correlations for individuals. The genotype correlations provided to individuals can be used to determine personal healthcare and lifestyle choices. If a strong correlation is found between a genetic variant and a treatable disease, detection of the genetic variant can aid in the decision to initiate disease treatment and/or individual monitoring. Where a statistically significant correlation exists but is not considered strong, the individual can discuss this information with a personal physician and decide on an appropriate, beneficial course of action. Potential courses of action that may benefit individuals with respect to specific genotype associations include therapeutic management, monitoring for potential treatment needs or effects, or lifestyle changes in terms of diet, exercise, and other personal habits/activities. For example, treatable phenotypes such as celiac disease could be treated symptomatically with a gluten-free diet. Also, through pharmacogenomics, genotype correlation information can be applied to predict the likely response of an individual who must be treated with a particular drug or course of drugs, such as the likely efficacy or safety of a particular drug treatment.

Registered users can choose to provide their genomic and phenotypic profiles to their healthcare managers, such as physicians or genetic counselors. The genomic and phenotypic profiles can be directly accessed by the healthcare manager, printed out by a registered user to give to the healthcare manager, or sent directly to the healthcare manager through an online portal (eg, via a link on an online report).

The delivery of this relevant information will allow the patient to take action in coordination with their physician. In particular, discussions between patients and their physicians can be enabled through personal portals and links to medical information and the integration of patients' genomic information into their medical records. Medical information can include preventive and health information. The information provided to individual patients by the present invention will enable patients to make informed choices about their healthcare. In this way, patients are able to make choices that help them avoid and/or delay diseases that are more likely to result from their individual genomic profile (inherited DNA). In addition, patients will be able to adopt treatment options tailored to their own specific medical needs. Individuals will also have the ability to access their genotype data should they develop a disease and need this information to help their doctors formulate treatment strategies.

Genotype association information can also be used in conjunction with genetic counseling to advise couples considering reproduction and to address potential genetic concerns for the mother, father, and/or child. A genetic counselor can provide information and support to registered users with a phenotypic profile that shows an increased risk of a particular condition or disease. They can interpret information about the condition, analyze inheritance patterns and risk of recurrence, and discuss available options with registered users. Genetic counselors can also provide supportive counseling and refer registered users to community or national support services. Genetic counseling can include certain registered programs. In some embodiments, genetic counseling can be scheduled within 24 hours of request and is available during hours such as evenings, Saturdays, Sundays, and/or holidays.

Individual entry will also facilitate the delivery of additional information beyond the initial screening. Individuals will be informed of new scientific discoveries about their personal genetic profile, such as information about new therapeutic or preventive strategies for their current or potential condition. New findings can also be passed on to their care managers. In a preferred embodiment, the registered user or his healthcare provider is notified by electronic mail of new genotype correlations and new studies regarding phenotypes in the registered user's phenotype profile. In other embodiments, emails of "recreational" phenotypes are sent to registered users, e.g. an e-letter may inform them that their genome profile is 77% identical to that of Abraham Lincoln and further information is available through an online portal supply.

The present invention also provides a method for generating new rules, amending rules, combining rules, regularly updating rule sets with new rules, securely maintaining a genome map database, applying rules to a genome map to determine phenotypic profiles and for generating reports computer code system. The computer code notifies registered users of new or revised correlations and new or revised reports, for example, reports with new prevention and health information, information about new treatments in development, or new treatments available.

business method

The present invention provides a commercial method for assessing an individual's genotype correlation based on a comparison of a patient's genomic profile to established clinical databases of medically relevant nucleotide variants. The present invention further provides a commercial method for evaluating initially unknown new correlations using a stored individual's genomic profile to generate an updated phenotypic profile of an individual without requiring the individual to submit additional biological samples. Figure 9 is a flowchart illustrating the business method.

The revenue stream of the business method of the present invention is generated in part in step 101 when an individual initially requests and purchases a personal genome profile for genotypic correlations of various common human diseases, conditions, and body states. Requests and purchases can be made through many sources including, but not limited to, online web portals, online health services, and an individual's personal physician or similar source of personal medical attention. In an alternative embodiment, the genomic profile may be provided free of charge, and a revenue stream may be generated in a subsequent step (eg, step 103).

A registered user or customer makes a request to purchase a phenotype profile. Collection kits are provided to consumers in response to demand and purchase for collection of the biological sample subjected to genetic sample isolation in step 103 . Provide collection kits by courier service, such as courier service offering same-day or overnight delivery, when requested online, by phone, or by other sources where collection kits are not readily available to consumers in person. Included in the collection kit are containers for the sample and packaging materials for rapid delivery of the sample to the laboratory where the genomic profile will be generated. The kit may also include instructions for sending the sample to a sample processing facility or laboratory and for accessing its genomic and phenotypic profiles, which may be done through an online portal.

As detailed above, genomic DNA can be obtained from any of a variety of types of biological samples. Preferably, genomic DNA is isolated from saliva using a commercially available collection kit (eg, from DNAGenotek). The use of saliva and this kit enables atraumatic sample collection as it is convenient for the consumer to provide a saliva sample in a container from the collection kit and then seal the container. Alternatively, saliva samples can be stored and shipped at room temperature.

After depositing the biological sample in a collection or specimen container, in step 105 the consumer delivers the sample to a laboratory for processing. Typically, the consumer can deliver/ship the sample to the laboratory using the packaging materials provided in the collection kit by expedited delivery, eg, same day or overnight courier service.

Laboratories that process samples and generate genomic profiles can follow appropriate government agency guidance and regulations. For example, in the United States, processing laboratories may be regulated by one or more federal agencies and/or one or more state agencies, such as the Food and Drug Administration (FDA) or the Centers for Medicare and Medicaid Services (CMS) . In the United States, clinical laboratories may be authorized or approved under the Clinical Laboratory Improvement Amendments (CLIA) of 1988.

In step 107, the sample is processed by the laboratory as previously described to isolate a genetic sample of DNA or RNA. Then, in step 109, the isolated genetic sample is analyzed and a genome map is generated. Preferably, a genomic SNP profile is generated. As mentioned above, SNP profiles can be generated using several methods. Preferably, high density arrays (such as commercially available platforms from Affymetrix or Illumina) are used for SNP identification and profile generation. For example, SNP profiles were generated using the AffymetrixGeneChipassay as described in more detail above. As technology develops, there may be other technology providers that can generate high-density SNP profiles. In another embodiment, the registered user's genome profile will be the registered user's genome sequence.

After the individual's genomic profile is generated, in step 111 the genotype data is preferably encrypted, entered, and in step 113 the data is stored in an encrypted database or vault where the information is stored for future use. The genomic profile and related information may be confidential, and access to this private information and genomic profile is limited as directed by the individual and/or his or her personal physician. Others (such as an individual's family and genetic counselors) may also be granted access by registered users.

The database or vault may be located locally at the processing laboratory. Alternatively, the database can be located at a separate location. In this case, the genomic profile data generated by the processing laboratory may be delivered in step 111 to a separate institution including a database.

After the individual's genomic profile is generated, the individual's genetic variation is then compared in step 115 to a clinical database of identified medically relevant genetic variants. Alternatively, a genotype correlation may not be medically relevant but still be included in a genotype correlation database, eg, a physical trait like eye color, or a "recreational" phenotype like similarity to a celebrity's genome profile.

Medically relevant SNPs can be established through scientific literature and related sources. Non-SNP genetic variants can also be established to correlate with phenotypes. Typically, SNP correlations for a given disease are established by comparing the haplotype patterns of a group of people known to have the disease to those without the disease. By analyzing many individuals, the frequency of a polymorphism in a population can be determined, and these genotype frequencies can then be correlated with a particular phenotype (eg, a disease or condition). Alternatively, the phenotype can be a non-medical state.

Associated SNP and non-SNP genetic variants can also be determined by analysis of stored individual genome profiles rather than from available published literature. Individuals with stored genomic profiles can reveal previously identified phenotypes. Analysis of an individual's genotype and revealed phenotype can be compared to individuals without the phenotype to determine correlations that can then be used for other genomic profiles. Individuals whose genomic profiles are determined may fill out questionnaires regarding previously identified phenotypes. The questionnaire can include questions about medical and non-medical conditions such as previously diagnosed diseases, family history of medical conditions, lifestyle, physical traits, mental traits, age, social life, environment, etc.

In one embodiment, individuals can have their genome profile determined for free if they fill out a questionnaire. In some embodiments, individuals periodically fill out questionnaires to gain free access to their phenotype profiles and reports. In other embodiments, individuals who complete the questionnaire can be given a registration upgrade so that they have a higher level of access than their previous registration, or they can purchase or renew their registration at a lower price.

In order to ensure scientific accuracy and significance, all information deposited in the medically relevant genetic variant database in step 121 is first approved by the research/clinical advisory group and, if authorized in step 119, by the appropriate government agencies Inspection and supervision. In the United States, for example, FDA can provide oversight by approving algorithms used to validate data associated with genetic variants (typically SNPs, transcript levels, or mutations). In step 123, the scientific literature and other relevant sources are monitored for additional genetic variant-disease or state correlations, and after confirming their accuracy and importance, and after review and approval by government agencies, these Additional genotype correlations are added to the master database in step 125 .

Databases of approved and validated medically relevant genetic variants combined with genome-wide individual profiles will advantageously allow genetic risk assessment for a large number of diseases or conditions. After compiling an individual's genomic profile, it can be done by comparing the individual's nucleotide (genetic) variants or genetic markers to a database of human nucleotide variants that have been associated with a particular phenotype (e.g., disease, condition, or body state) while determining individual genotype correlations. By comparing an individual's genomic profile to a master database of genotype correlations, individuals can be informed whether and to what extent they are found to be positive or negative for genetic risk factors. Individuals will receive relative risk and/or predisposition data for a wide range of scientifically validated disease states (eg, Alzheimer's disease, cardiovascular disease, blood coagulation). For example, the genotype correlations in Table 1 can be included. Additionally, SNP disease correlations in the database may include, but are not limited to, those correlations shown in FIG. 4 . Other dependencies in FIGS. 5 and 6 may also be included. The business methods of the present invention thus provide risk analysis for a large number of diseases and conditions without prior knowledge of what risks those diseases and conditions may cause.

In other embodiments, the genotype correlations combined with the genome-wide individual profile are non-medically relevant phenotypes, such as "recreational" phenotypes or physical traits such as hair color. In a preferred embodiment, a rule or set of rules is applied to an individual's genomic profile or SNP profile, as described above. Applying the rules to the genomic profile generates a phenotypic profile for the individual.

Thus, the master database of human genotype correlations is expanded with additional genotype correlations as new correlations are discovered and validated. When needed or appropriate, it can be updated by accessing relevant information from the individual's genome profile stored in the database. For example, learned novel genotype associations may be based on specific genetic variants. Whether an individual is likely to be affected by this new genotype correlation can then be determined by obtaining and comparing only that gene's portion of the individual's complete genome profile.

The results of the genomic query are preferably analyzed and interpreted for presentation to the individual in an understandable form. Then, in step 117, the results of the initial screening are provided to the patient in a secure, confidential manner by mail or through an online portal interface as described in detail above.

A report may include a phenotype profile as well as genomic information about the phenotypes in the phenotype profile, for example, basic genetic information about the genes involved or statistical information about genetic variants in different populations. Other information based on phenotype profiles that can be included in the report are preventive strategies, health information, treatment methods, symptom awareness, early detection protocols, intervention protocols, and further identification and classification of phenotypes. After the initial screening of an individual's genomic profile, controlled, modest updates are or can be made.

Updates to the individual's genome profile are made or available in conjunction with updates to the master database as new genotype correlations arise and are validated and approved. New rules based on new genotype correlations can be applied to the initial genomic profile to provide an updated phenotypic profile. An updated genotype correlation profile may be generated in step 127 by comparing the relevant portion of the individual's genomic profile with the new genotype correlation. For example, if a new genotype association is discovered based on a variation in a particular gene, that gene portion of the individual's genomic profile can be analyzed for the new genotype association. In this case, one or more rules may be applied to generate an updated phenotype profile, rather than updating the phenotype profile with the entire rule set with rules already applied. In step 129, the results of the individual's updated genotype correlations are provided in encrypted form.

Initial and updated phenotype profiles may be a service provided to registered users or customers. Different registration levels and combinations thereof for genomic profiling are available. Likewise, enrollment levels can be varied to provide individuals with a choice of the amount of services they wish to receive in relation to their genotype. In this way, the level of service offered will vary with the level of service registration an individual purchases.

An entry-level registration for a registered user can include a genomic profile and an initial phenotypic profile. This can be the base registration level. There can be different service levels within the base registration level. For example, specific registration levels may provide access to genetic counseling, physicians with particular expertise in treating or preventing specific diseases, and other service options. Genetic counseling is available online or by phone. In another embodiment, the price of registration may depend on the number of phenotypes an individual chooses for their phenotype profile. Another option might be whether registered users choose to access online genetic counseling.

In another instance, a registry can provide an initial genome-wide genotype correlation while maintaining an individual's genomic profile in a database; this database can be encrypted if the individual so chooses. After this initial analysis, subsequent analysis and additional results can be done upon individual request and additional payment. This can be a premium registration.

In one embodiment of the business method according to the invention, an update of the individual's risk is carried out and corresponding information can be provided to the individual on the basis of registration. Updates are available to registered users who purchase premium registrations. Registration for genotype correlation analysis can provide updates for specific types or subclasses of new genotype correlations according to individual preferences. For example, an individual may only wish to be informed of genotype correlations for which there is a known course of treatment or prevention. To assist an individual in deciding whether to perform additional analyses, the individual can be provided with information about additional genotype correlations that are already available. This information can be conveniently mailed or emailed to registered users.

In advanced registration there can be more service levels such as those mentioned in basic registration. Other registration modes may be provided at a high level. For example, the highest tier may provide registered users with unlimited updates and reports. The profile of registered users may be updated as new correlations and rules are determined. At this level, registered users can also allow access to an unlimited number of individuals, such as family members and healthcare managers. Registered users also get unlimited access to online genetic counselors and doctors.

The next level of registration within a high level may offer more restricted aspects, such as a limited number of updates. Registered users can update their genome profile a limited number of times during the registration period, for example, 4 times a year. In another level of registration, registered users can update their stored genomic profiles once a week, once a month, or once a year. In another embodiment, registered users may only have a limited number of phenotypes that they may choose to update their genomic profile.

The personal portal will also conveniently enable individuals to maintain registration for risk or relevance updates and/or information updates, or to request updated risk assessments and information. As noted above, different enrollment levels may be provided to enable individuals to select various levels of genotype correlation results and updates, and registered users may select different enrollment levels through their personal portal.

Any of these registration options will contribute to the revenue stream of the present invention business method. The revenue stream of the business method of the present invention is also increased by adding new customers and registered users, where new genome profiles are added to the database.

Table 1: Typical genes with genetic variants associated with phenotypes.

Gene Phenotype A2M Alzheimer's disease ABCA1 Cholesterol, HDL ABCB1 HIV ABCB1 epilepsy ABCB1 Kidney Transplant Complications ABCB1 digoxin, serum concentration ABCB1 Crohn's disease; ulcerative colitis ABCB1 Parkinson's disease ABCC8 type 2 diabetes ABCC8 diabetes, type 2 ABOs myocardial infarction ACADM Medium-chain acyl-CoA dehydrogenase deficiency ACDC type 2, diabetes ACE type 2 diabetes ACE hypertension ACE Alzheimer's disease ACE myocardial infarction ACE cardiovascular ACE left ventricular hypertrophy

Gene Phenotype ACE coronary artery disease ACE atherosclerosis, coronary ACE retinopathy, diabetic ACE systemic lupus erythematosus ACE blood pressure, arterial ACE erectile dysfunction ACE lupus ACE polycystic kidney disease ACE stroke ACP1 diabetes, type 1 ACSM1(LIP)c cholesterol level ADAM33 asthma ADD1 hypertension ADD1 blood pressure, arterial ADH1B alcohol abuse ADH1C alcohol abuse ADIPOQ diabetes, type 2 ADIPOQ obesity ADORA2A panic attack ADRB1 hypertension ADRB1 heart failure ADRB2 asthma ADRB2 hypertension ADRB2 obesity ADRB2 blood pressure, arterial ADRB2 type 2 diabetes ADRB3 obesity

Gene Phenotype ADRB3 type 2 diabetes ADRB3 hypertension AGT hypertension AGT type 2 diabetes AGT essential hypertension AGT myocardial infarction AGTR1 hypertension AGTR2 hypertension AHR breast cancer ALAD lead toxicity ALDH2 alcoholism ALDH2 alcohol abuse ALDH2 colorectal cancer ALDRL2 type 2 diabetes ALOX5 asthma ALOX5AP asthma APBB1 Alzheimer's disease APCs colorectal cancer APEX1 lung cancer APOA1 Atherosclerosis, coronary APOA1 Cholesterol, HDL APOA1 coronary artery disease APOA1 type 2 diabetes APOA4 type 2 diabetes APOA5 Triglycerides APOA5 Atherosclerosis, coronary APOB Hypercholesterolemia

Gene Phenotype APOB obesity APOB cardiovascular APOB coronary artery disease APOB coronary heart disease APOB type 2 diabetes APOC1 Alzheimer's disease APOC3 Triglycerides APOC3 type 2 diabetes APOE Alzheimer's disease APOE type 2 diabetes APOE multiple sclerosis APOE Atherosclerosis, coronary APOE Parkinson's disease APOE coronary heart disease APOE myocardial infarction APOE stroke APOE Alzheimer's disease APOE coronary artery disease app Alzheimer's disease AR prostate cancer AR breast cancer ATMs breast cancer ATP7B Wilson disease ATXN8OS spinocerebellar ataxia BACE1 Alzheimer's disease BCHE Alzheimer's disease BDKRB2 hypertension

Gene Phenotype BDNF Alzheimer's disease BDNF bipolar disorder BDNF Parkinson's disease BDNF schizophrenia BDNF memory BGLAP bone density BRAF Thyroid cancer BRCA1 breast cancer BRCA1 breast cancer; ovarian cancer BRCA1 ovarian cancer BRCA2 breast cancer BRCA2 breast cancer; ovarian cancer BRCA2 ovarian cancer BRIP1 breast cancer C4A systemic lupus erythematosus CALCR bone density CAMTA1 episodic memory CAPN10 diabetes, type 2 CAPN10 type 2 diabetes CAPN3 muscular dystrophy CARD15 Crohn's disease CARD15 Crohn's disease; ulcerative colitis CARD15 Inflammatory bowel disease CART obesity CASR bone density CCKAR schizophrenia CCL2 systemic lupus erythematosus

Gene Phenotype CCL5 HIV CCL5 asthma CCND1 colorectal cancer CCR2 HIV CCR2 HIV infection CCR2 Hepatitis C CCR2 myocardial infarction CCR3 asthma CCR5 HIV CCR5 HIV infection CCR5 Hepatitis C CCR5 asthma CCR5 multiple sclerosis CD14 Atopy CD14 asthma CD14 Crohn's disease CD14 Crohn's disease; ulcerative colitis CD14 Periodontitis CD14 total IgE CDH1 prostate cancer CDH1 colorectal cancer CDKN2A melanoma CDSN psoriasis CEBPA leukemia, bone marrow CETP Atherosclerosis, coronary CETP coronary heart disease CETP Hypercholesterolemia

Gene Phenotype CFH macular degeneration CFTR cystic fibrosis CFTR pancreatitis CFTR cystic fibrosis CHAT Alzheimer's disease CHEK2 breast cancer CHRNA7 schizophrenia CMA1 atopic dermatitis CNR1 schizophrenia COL1A1 bone density COL1A1 osteoporosis COL1A2 bone density COL2A1 Osteoarthritis COMT schizophrenia COMT breast cancer COMT Parkinson's disease COMT bipolar disorder COMT obsessive-compulsive neurosis COMT alcoholism CR1 systemic lupus erythematosus CRP C-reactive protein CST3 Alzheimer's disease CTLA4 type 1 diabetes CTLA4 Graves' disease CTLA4 multiple sclerosis CTLA4 rheumatoid arthritis CTLA4 systemic lupus erythematosus

Gene Phenotype CTLA4 lupus erythematosus CTLA4 celiac disease CTSD Alzheimer's disease CX3CR1 HIV CXCL12 HIV CXCL12 HIV infection CYBA Atherosclerosis, coronary CYBA hypertension CYP11B2 hypertension CYP11B2 left ventricular hypertrophy CYP17A1 breast cancer CYP17A1 prostate cancer CYP17A1 endometriosis CYP17A1 endometrial cancer CYP19A1 breast cancer CYP19A1 prostate cancer CYP19A1 endometriosis CYP1A1 lung cancer CYP1A1 breast cancer CYP1A1 colorectal cancer CYP1A1 prostate cancer CYP1A1 Esophageal cancer CYP1A1 endometriosis CYP1A1 Cytogenesis Research CYP1A2 schizophrenia CYP1A2 colorectal cancer CYP1B1 breast cancer

Gene Phenotype CYP1B1 glaucoma CYP1B1 prostate cancer CYP21A2 21-hydroxylase deficiency CYP21A2 congenital adrenal hyperplasia CYP21A2 Adrenal hyperplasia, congenital CYP2A6 smoking behavior CYP2A6 Nicotine CYP2A6 lung cancer CYP2C19 Helicobacter pylori infection CYP2C19 Phenytoin CYP2C19 stomach trouble CYP2C8 Malaria, Plasmodium falciparum CYP2C9 Anticoagulant Complications CYP2C9 Warren Sensitivity CYP2C9 warfarin therapy, its response CYP2C9 colorectal cancer CYP2C9 Phenytoin CYP2C9 Acenocoumarol reaction CYP2C9 coagulation disorder CYP2C9 hypertension CYP2D6 colorectal cancer CYP2D6 Parkinson's disease CYP2D6 CYP2D6 poor metabolizer phenotype CYP2E1 lung cancer CYP2E1 colorectal cancer CYP3A4 prostate cancer CYP3A5 prostate cancer

Gene Phenotype CYP3A5 Esophageal cancer CYP46A1 Alzheimer's disease DBH schizophrenia DHCR7 Steiner-Osman Syndrome DISC1 schizophrenia DLST Alzheimer's disease DMD muscular dystrophy DRD2 alcoholism DRD2 schizophrenia DRD2 smoking behavior DRD2 Parkinson's disease DRD2 tardive dyskinesia DRD3 schizophrenia DRD3 tardive dyskinesia DRD3 bipolar disorder DRD4 attention deficit disorder [with hyperactivity] DRD4 schizophrenia DRD4 novelty seeking DRD4 ADHD DRD4 personality qualities DRD4 heroin abuse DRD4 alcohol abuse DRD4 alcoholism DRD4 personality disorder DTNBP1 schizophrenia EDN1 hypertension EGFR lung cancer

Gene Phenotype ELAC2 prostate cancer ENPP1 type 2 diabetes EPHB2 prostate cancer EPHX1 lung cancer EPHX1 colorectal cancer EPHX1 Cell Generation Research EPHX1 Chronic Obstructive Pulmonary Disease/COPD ERBB2 breast cancer ERCC1 lung cancer ERCC1 colorectal cancer ERCC2 lung cancer ERCC2 Cell Generation Research ERCC2 Bladder Cancer ERCC2 colorectal cancer ESR1 bone density ESR1 bone mineral density ESR1 breast cancer ESR1 endometriosis ESR1 osteoporosis ESR2 bone density ESR2 breast cancer estrogen receptor bone mineral density F2 coronary heart disease F2 stroke F2 Thromboembolism, Venous F2 preeclampsia F2 thrombosis

Gene Phenotype F5 Thromboembolism, Venous F5 preeclampsia F5 myocardial infarction F5 stroke F5 stroke, ischemic F7 Atherosclerosis, coronary F7 myocardial infarction F8 hemophilia F9 hemophilia FABP2 type 2 diabetes FAS Alzheimer's disease FASLG multiple sclerosis FCGR2A systemic lupus erythematosus FCGR2A lupus erythematosus FCGR2A Periodontitis FCGR2A rheumatoid arthritis FCGR2B lupus erythematosus FCGR2B systemic lupus erythematosus FCGR3A systemic lupus erythematosus FCGR3A lupus erythematosus FCGR3A Periodontitis FCGR3A arthritis FCGR3A rheumatoid arthritis FCGR3B Periodontitis FCGR3B periodontal disease FCGR3B lupus erythematosus FGB Fibrinogen

Gene Phenotype FGB myocardial infarction FGB coronary heart disease FLT3 leukemia, bone marrow FLT3 leukemia FMR1 fragile X syndrome FRAXA fragile X syndrome FUT2 Helicobacter pylori infection FVL Factor V Leiden G6PD G6PD deletion G6PD hyperbilirubinemia GABRA5 bipolar disorder GBA Gaucher disease GBA Parkinson's disease GCGR (FAAH, ML4R, UCP2) weight/obesity GCK type 2 diabetes GCLM (F12, TLR4) Atherosclerosis, myocardial infarction GDNF schizophrenia GHRL obesity GJB1 Charcot-Marie-Tooth disease GJB2 deaf GJB2 Hearing loss, sensorineural nonsyndromic GJB2 hearing loss, sensorineural GJB2 hearing loss/deafness GJB6 Hearing loss, sensorineural nonsyndromic GJB6 hearing loss/deafness GNAS hypertension GNB3 hypertension

Gene Phenotype GPX1 lung cancer GRIN1 schizophrenia GRIN2B schizophrenia GSK3B bipolar disorder GSTM1 lung cancer GSTM1 colorectal cancer GSTM1 breast cancer GSTM1 prostate cancer GSTM1 Cell Generation Research GSTM1 Bladder Cancer GSTM1 Esophageal cancer GSTM1 head and neck cancer GSTM1 leukemia GSTM1 Parkinson's disease GSTM1 stomach cancer GSTP1 lung cancer GSTP1 colorectal cancer GSTP1 breast cancer GSTP1 Cell Generation Research GSTP1 prostate cancer GSTT1 lung cancer GSTT1 colorectal cancer GSTT1 breast cancer GSTT1 prostate cancer GSTT1 Bladder Cancer GSTT1 Cell Generation Research GSTT1 asthma

Gene Phenotype GSTT1 Benzene toxicity GSTT1 Esophageal cancer GSTT1 head and neck cancer GYS1 type 2 diabetes HBB Thalassemia HBB Thalassemia, beta- HD Huntington's disease HFE Hemochromatosis HFE iron level HFE colorectal cancer HK2 type 2 diabetes HLA rheumatoid arthritis HLA type 1 diabetes HLA Behcet's disease HLA celiac disease HLA psoriasis HLA Graves' disease HLA multiple sclerosis HLA schizophrenia HLA asthma HLA diabetes HLA lupus HLA-A leukemia HLA-A HIV HLA-A diabetes, type 1 HLA-A graft versus host disease HLA-A multiple sclerosis

Gene Phenotype HLA-B leukemia HLA-B Behcet's disease HLA-B celiac disease HLA-B diabetes, type 1 HLA-B graft versus host disease HLA-B Sarcoidosis HLA-C psoriasis HLA-DPA1 measles HLA—DPB1 diabetes, type 1 HLA—DPB1 asthma HLA-DQA1 diabetes, type 1 HLA-DQA1 celiac disease HLA-DQA1 cervical cancer HLA-DQA1 asthma HLA-DQA1 multiple sclerosis HLA-DQA1 Diabetes mellitus, type 2; Diabetes mellitus, type 1 HLA-DQA1 lupus erythematosus HLA-DQA1 pregnancy loss, recurrent HLA-DQA1 psoriasis HLA-DQB1 diabetes, type 1 HLA-DQB1 celiac disease HLA-DQB1 multiple sclerosis HLA-DQB1 cervical cancer HLA-DQB1 lupus erythematosus HLA-DQB1 pregnancy loss, recurrent HLA-DQB1 arthritis HLA-DQB1 asthma

Gene Phenotype HLA-DQB1 HIV HLA-DQB1 Lymphoma HLA-DQB1 tuberculosis HLA-DQB1 rheumatoid arthritis HLA-DQB1 diabetes, type 2 HLA-DQB1 graft versus host disease HLA-DQB1 Narcolepsy HLA-DQB1 arthritis, rheumatoid HLA-DQB1 Cholangitis, sclerosing HLA-DQB1 Diabetes mellitus, type 2; Diabetes mellitus, type 1 HLA-DQB1 Graves' disease HLA-DQB1 Hepatitis C HLA-DQB1 hepatitis C, chronic HLA-DQB1 malaria HLA-DQB1 Malaria, Plasmodium falciparum HLA-DQB1 melanoma HLA-DQB1 psoriasis HLA-DQB1 Sjögren syndrome HLA-DQB1 systemic lupus erythematosus HLA-DRB1 diabetes, type 1 HLA-DRB1 multiple sclerosis HLA-DRB1 systemic lupus erythematosus HLA-DRB1 rheumatoid arthritis HLA-DRB1 cervical cancer HLA-DRB1 arthritis HLA-DRB1 celiac disease HLA-DRB1 lupus erythematosus

Gene Phenotype HLA-DRB1 Sarcoidosis HLA-DRB1 HIV HLA-DRB1 tuberculosis HLA-DRB1 Graves' disease HLA-DRB1 Lymphoma HLA-DRB1 psoriasis HLA-DRB1 asthma HLA-DRB1 Crohn's disease HLA-DRB1 graft versus host disease HLA-DRB1 hepatitis C, chronic HLA-DRB1 Narcolepsy HLA-DRB1 sclerosis, generalized HLA-DRB1 Sjögren syndrome HLA-DRB1 type 1 diabetes HLA-DRB1 arthritis, rheumatoid HLA-DRB1 Cholangitis, sclerosing HLA-DRB1 Diabetes mellitus, type 2; Diabetes mellitus, type 1 HLA-DRB1 Helicobacter pylori infection HLA-DRB1 Hepatitis C HLA-DRB1 juvenile arthritis HLA-DRB1 leukemia HLA-DRB1 malaria HLA-DRB1 melanoma HLA-DRB1 pregnancy loss, recurrent HLA-DRB3 psoriasis HLA-G pregnancy loss, recurrent HMOX1 Atherosclerosis, coronary

Gene Phenotype HNF4A diabetes, type 2 HNF4A type 2 diabetes HSD11B2 hypertension HSD17B1 breast cancer HTR1A depression, severe HTR1B alcohol dependence HTR1B alcoholism HTR2A memory HTR2A schizophrenia HTR2A bipolar disorder HTR2A depression HTR2A depression, severe HTR2A suicide HTR2A Alzheimer's disease HTR2A anorexia nervosa HTR2A hypertension HTR2A obsessive-compulsive neurosis HTR2C schizophrenia HTR6 Alzheimer's disease HTR6 schizophrenia HTRA1 wet age-related macular degeneration IAPP type 2 diabetes IDEs Alzheimer's disease IFNG tuberculosis IFNG type 1 diabetes IFNG graft versus host disease IFNG Hepatitis B

Gene Phenotype IFNG multiple sclerosis IFNG asthma IFNG breast cancer IFNG Kidney transplant IFNG Kidney Transplant Complications IFNG longevity IFNG pregnancy loss, recurrent IGFBP3 breast cancer IGFBP3 prostate cancer IL10 systemic lupus erythematosus IL10 asthma IL10 graft versus host disease IL10 HIV IL10 Kidney transplant IL10 Kidney Transplant Complications IL10 Hepatitis B IL10 juvenile arthritis IL10 longevity IL10 multiple sclerosis IL10 pregnancy loss, recurrent IL10 rheumatoid arthritis IL10 tuberculosis IL12B type 1 diabetes IL12B asthma IL13 asthma IL13 Atopy IL13 Chronic Obstructive Pulmonary Disease/COPD

Gene Phenotype IL13 Graves' disease IL1A Periodontitis IL1A Alzheimer's disease IL1B Periodontitis IL1B Alzheimer's disease IL1B stomach cancer IL1R1 type 1 diabetes IL1RN stomach cancer IL2 Asthma; Eczema; Allergic diseases IL4 asthma IL4 Atopy IL4 HIV IL4R asthma IL4R Atopy IL4R total serum IgE IL6 bone mineralization IL6 Kidney transplant IL6 Kidney Transplant Complications IL6 longevity IL6 multiple sclerosis IL6 bone density IL6 bone mineral density IL6 colorectal cancer IL6 juvenile arthritis IL6 rheumatoid arthritis IL9 asthma INHA premature ovarian failure

Gene Phenotype INS type 1 diabetes INS type 2 diabetes INS diabetes, type 1 INS obesity INS prostate cancer INSIG2 obesity INSR type 2 diabetes INSR hypertension INSR polycystic ovary syndrome IPF1 diabetes, type 2 IRS1 type 2 diabetes IRS1 diabetes, type 2 IRS2 diabetes, type 2 ITGB3 myocardial infarction ITGB3 Atherosclerosis, coronary ITGB3 coronary heart disease ITGB3 myocardial infarction KCNE1 EKG, abnormal KCNE2 EKG, abnormal KCNH2 EKG, abnormal KCNH2 long QT syndrome KCNJ11 diabetes, type 2 KCNJ11 type 2 diabetes KCNN3 schizophrenia KCNQ1 EKG, abnormal KCNQ1 long QT syndrome KIBRA episodic memory

Gene Phenotype KLK1 hypertension KLK3 prostate cancer KRAS colorectal cancer LDLR Hypercholesterolemia LDLR hypertension LEP obesity LEPR obesity LIG4 breast cancer LIPCs Atherosclerosis, coronary LPL coronary artery disease LPL Hyperlipidemia LPL Triglycerides LRP1 Alzheimer's disease LRP5 bone density LRRK2 Parkinson's disease LRRK2 Parkinson's Disease LTA type 1 diabetes LTA asthma LTA systemic lupus erythematosus LTA septicemia LTC4S asthma MAOA alcoholism MAOA schizophrenia MAOA bipolar disorder MAOA smoking behavior MAOA personality disorder MAOB Parkinson's disease

Gene Phenotype MAOB smoking behavior MAPT Parkinson's disease MAPT Alzheimer's disease MAPT dementia MAPT frontotemporal dementia MAPT progressive supranuclear palsy MC1R melanoma MC3R obesity MC4R obesity MECP2 Rett syndrome MEFV familial mediterranean fever MEFV Amyloidosis MICA type 1 diabetes MICA Behcet's disease MICA celiac disease MICA rheumatoid arthritis MICA systemic lupus erythematosus MLH1 colorectal cancer MME Alzheimer's disease MMP1 lung cancer MMP1 ovarian cancer MMP1 Periodontitis MMP3 myocardial infarction MMP3 ovarian cancer MMP3 rheumatoid arthritis MPO lung cancer MPO Alzheimer's disease

Gene Phenotype MPO breast cancer MPZ Charcot-Marie-Tooth disease MS4A2 asthma MS4A2 Atopy MSH2 colorectal cancer MSH6 colorectal cancer MSR1 prostate cancer MTHFR colorectal cancer MTHFR type 2 diabetes MTHFR neural tube defect MTHFR homocysteine MTHFR Thromboembolism, Venous MTHFR Atherosclerosis, coronary MTHFR Alzheimer's disease MTHFR Esophageal cancer MTHFR preeclampsia MTHFR pregnancy loss, recurrent MTHFR stroke MTHFR thrombosis, deep vein MT—ND1 diabetes, type 2 MTR colorectal cancer MT—RNR1 Hearing loss, sensorineural non-syndromic MTRR neural tube defect MTRR homocysteine MT—TL1 diabetes, type 2 MUTYH colorectal cancer MYBPC3 Cardiomyopathy

Gene Phenotype MYH7 Cardiomyopathy MYOC Glaucoma, open angle primary MYOC glaucoma NAT1 colorectal cancer NAT1 breast cancer NAT1 Bladder Cancer NAT2 colorectal cancer NAT2 Bladder Cancer NAT2 breast cancer NAT2 lung cancer NBN breast cancer NCOA3 breast cancer NCSTN Alzheimer's disease NEUROD1 type 1 diabetes NF1 Neurofibromatosis 1 NOS1 asthma NOS2A multiple sclerosis NOS3 hypertension NOS3 coronary heart disease NOS3 Atherosclerosis, coronary NOS3 coronary artery disease NOS3 myocardial infarction NOS3 acute coronary syndrome NOS3 blood pressure, arterial NOS3 preeclampsia NOS3 Nitric oxide NOS3 Alzheimer's disease

Gene Phenotype NOS3 asthma NOS3 type 2 diabetes NOS3 Cardiovascular disease NOS3 Behcet's disease NOS3 erectile dysfunction NOS3 kidney failure, chronic NOS3 lead toxicity NOS3 left ventricular hypertrophy NOS3 pregnancy loss, recurrent NOS3 retinopathy, diabetic NOS3 stroke NOTCH4 schizophrenia NPY alcohol abuse NQO1 lung cancer NQO1 colorectal cancer NQO1 Benzene toxicity NQO1 Bladder Cancer NQO1 Parkinson's disease NR3C2 hypertension NR4A2 Parkinson's disease NRG1 schizophrenia NTF3 schizophrenia OGG1 lung cancer OGG1 colorectal cancer OLR1 Alzheimer's disease OPA1 glaucoma OPRM1 alcohol abuse

Gene Phenotype OPRM1 drug dependence OPTN Glaucoma, open angle primary P450 drug metabolism PADI4 rheumatoid arthritis PAH Phenylketonuria/PKU PAI1 coronary heart disease PAI1 asthma PALB2 breast cancer PARK2 Parkinson's disease PARK7 Parkinson's disease PDCD1 lupus erythematosus PINK1 Parkinson's disease PKA memory PKC memory PLA2G4A schizophrenia PNOC schizophrenia POMC obesity PON1 Atherosclerosis, coronary PON1 Parkinson's disease PON1 type 2 diabetes PON1 atherosclerosis PON1 coronary artery disease PON1 coronary heart disease PON1 Alzheimer's disease PON1 longevity PON2 Atherosclerosis, coronary PON2 premature birth

Gene Phenotype PPARG type 2 diabetes PPARG obesity PPARG diabetes, type 2 PPARG colorectal cancer PPARG hypertension PPARGC1A diabetes, type 2 PRKCZ type 2 diabetes PRL systemic lupus erythematosus PRNP Alzheimer's disease PRNP Creutzfeldt-Jakob disease PRNP Jacques-Craft disease PRODH schizophrenia PRSS1 pancreatitis PSEN1 Alzheimer's disease PSEN2 Alzheimer's disease PSMB8 type 1 diabetes PSMB9 type 1 diabetes PTCH Skin cancer, non-melanoma PTGIS hypertension PTGS2 colorectal cancer PTH bone density PTPN11 Noonan Syndrome PTPN22 rheumatoid arthritis PTPRC multiple sclerosis PVT1 end stage renal disease RAD51 breast cancer RAGE retinopathy, diabetic

Gene Phenotype RB1 Retinoblastoma RELN schizophrenia REN hypertension RET Thyroid cancer RET Hirschsprung's disease RFC1 neural tube defect RGS4 schizophrenia RHO retinitis pigmentosa RNASEL prostate cancer RYR1 malignant hyperthermia SAA1 Amyloidosis SCG2 hypertension SCG3 obesity SCGB1A1 asthma SCN5A Brugada syndrome SCN5A EKG, abnormal SCN5A long QT syndrome SCNN1B hypertension SCNN1G hypertension SERPINA1 COPD SERPINA3 Alzheimer's disease SERPINA3 COPD SERPINA3 Parkinson's disease SERPINE1 myocardial infarction SERPINE1 type 2 diabetes SERPINE1 Atherosclerosis, coronary SERPINE1 obesity

Gene Phenotype SERPINE1 preeclampsia SERPINE1 stroke SERPINE1 hypertension SERPINE1 pregnancy loss, recurrent SERPINE1 Thromboembolism, Venous SLC11A1 tuberculosis SLC22A4 Crohn's disease; ulcerative colitis SLC22A5 Crohn's disease; ulcerative colitis SLC2A1 type 2 diabetes SLC2A2 type 2 diabetes SLC2A4 type 2 diabetes SLC3A1 cystinuria SLC6A3 attention deficit disorder [with hyperactivity] SLC6A3 Parkinson's disease SLC6A3 smoking behavior SLC6A3 alcoholism SLC6A3 schizophrenia SLC6A4 depression SLC6A4 depression, severe SLC6A4 schizophrenia SLC6A4 suicide SLC6A4 alcoholism SLC6A4 bipolar disorder SLC6A4 personality qualities SLC6A4 attention deficit disorder [with hyperactivity] SLC6A4 Alzheimer's disease SLC6A4 personality disorder

Gene Phenotype SLC6A4 panic attack SLC6A4 alcohol abuse SLC6A4 affective disorder SLC6A4 anxiety disorder SLC6A4 smoking behavior SLC6A4 Depression, severe; bipolar disorder SLC6A4 heroin abuse SLC6A4 irritable bowel syndrome SLC6A4 Migraine SLC6A4 obsessive-compulsive neurosis SLC6A4 suicidal behavior SLC7A9 cystinuria SNAP25 ADHD SNCA Parkinson's disease SOD1 ALS/Amyotrophic Lateral Sclerosis SOD2 breast cancer SOD2 lung cancer SOD2 prostate cancer SPINK1 pancreatitis SPP1 multiple sclerosis SRD5A2 prostate cancer STAT6 asthma STAT6 total IgE SULT1A1 breast cancer SULT1A1 colorectal cancer TAP1 type 1 diabetes TAP1 lupus erythematosus

Gene Phenotype TAP2 type 1 diabetes TAP2 diabetes, type 1 TBX21 asthma TBXA2R asthma TCF1 diabetes, type 2 TCF1 type 2 diabetes TF Alzheimer's disease TGFB1 breast cancer TGFB1 Kidney transplant TGFB1 Kidney Transplant Complications TH schizophrenia THBD myocardial infarction TLR4 asthma TLR4 Crohn's disease; ulcerative colitis TLR4 septicemia TNF asthma TNFA Cerebrovascular disease TNF type 1 diabetes TNF rheumatoid arthritis TNF systemic lupus erythematosus TNF Kidney transplant TNF psoriasis TNF septicemia TNF type 2 diabetes TNF Alzheimer's disease TNF Crohn's disease TNF diabetes, type 1

Gene Phenotype TNF Hepatitis B TNF Kidney Transplant Complications TNF multiple sclerosis TNF schizophrenia TNF celiac disease TNF obesity TNF pregnancy loss, recurrent TNFRSF11B bone density TNFRSF1A rheumatoid arthritis TNFRSF1B rheumatoid arthritis TNFRSF1B systemic lupus erythematosus TNFRSF1B arthritis TNNT2 Cardiomyopathy TP53 lung cancer TP53 breast cancer TP53 colorectal cancer TP53 prostate cancer TP53 cervical cancer TP53 ovarian cancer TP53 smoking TP53 Esophageal cancer TP73 lung cancer TPH1 suicide TPH1 depression, severe TPH1 suicidal behavior TPH1 schizophrenia TPMT Thiopurine methyltransferase activity

Gene Phenotype TPMT leukemia TPMT Inflammatory bowel disease TPMT Thiopurine S-methyltransferase Phenotype TSC1 tuberous sclerosis TSC2 tuberous sclerosis TSHR Graves' disease TYMS colorectal cancer TYMS stomach cancer TYMS Esophageal cancer UCHL1 Parkinson's disease UCP1 obesity UCP2 obesity UCP3 obesity UGT1A1 hyperbilirubinemia UGT1A1 Gerber syndrome UGT1A6 colorectal cancer UGT1A7 colorectal cancer UTS2 diabetes, type 2 VDR bone density VDR prostate cancer VDR bone mineral density VDR type 1 diabetes VDR osteoporosis VDR bone mass VDR breast cancer VDR lead toxicity VDR tuberculosis

Gene Phenotype VDR type 2 diabetes VEGF breast cancer Vit D rec idiopathic short stature VKORC1 warfarin therapy, its response WNK4 hypertension XPA lung cancer XPC lung cancer XPC Cell Generation Research XRCC1 lung cancer XRCC1 Cell Generation Research XRCC1 breast cancer XRCC1 Bladder Cancer XRCC2 breast cancer XRCC3 breast cancer XRCC3 Cell Generation Research XRCC3 lung cancer XRCC3 Bladder Cancer ZDHHC8 schizophrenia

Genetic Composite Index (GCI)

The etiology of many states or diseases has been attributed to genetic and environmental factors. Recent advances in genotyping technology have provided opportunities to identify new associations between diseases and genetic markers across the genome. Indeed, many recent studies have found these associations, in which specific alleles or genotypes are associated with increased disease risk. Some of these studies involve collecting a set of test cases and a set of controls and comparing the allelic distribution of genetic markers between the two populations. In some of these studies, the association between a particular genetic marker and disease was determined in isolation from other genetic markers, which were treated as background and played no role in the statistical analysis.

Genetic markers and variants may include SNPs, nucleotide duplications, nucleotide insertions, nucleotide deletions, chromosomal translocations, chromosomal duplications, or copy number variations. Copy number variations may include microsatellite repeats, nucleotide repeats, centromeric repeats, or telomeric repeats.

In one aspect of the invention, information on the association of multiple genetic markers with one or more diseases or conditions is combined and analyzed to obtain a GCI score. The GCI score can be used to provide individuals with no genetics training a reliable (ie, robust), comprehensible and/or intuitive view of their individual risk of disease compared to a relevant population based on current scientific research. In one embodiment, the method of generating a reliable GCI score for the combined effect of different loci is based on the reported individual risk for each studied locus. For example, a disease or condition of interest is identified, and information sources (including, but not limited to, databases, patent publications, and scientific literature) are consulted for information on the association of the disease or condition with one or more genetic loci. These sources of information are verified and evaluated using quality criteria. In some embodiments, the assessment process includes multiple steps. In other embodiments, information sources are evaluated on multiple quality criteria. Information derived from information sources is used to identify odds ratios or relative risks for one or more genetic loci for each disease or condition of interest.

In alternative embodiments, the odds ratio (OR) or relative risk (RR) for at least one genetic locus cannot be obtained from available sources of information. Then use (1) reported ORs of multiple alleles for the same locus, (2) allele frequencies from a dataset (e.g., HapMap dataset), and/or (3) from available resources (e.g., CDC, National Center for Health Statistics, etc.) calculates RR for disease/state prevalence to derive RR for all alleles of interest. In one embodiment, ORs for multiple alleles of the same locus are assessed separately or independently. In a preferred embodiment, ORs for multiple alleles of the same locus are combined to account for dependencies between ORs for different alleles. In some embodiments, established disease models (including, but not limited to, multiplicative, additive, Harvard-modified, dominant-effects models) are used to generate a model representing individual risk according to the selected model. Intermediate rating.

In another embodiment, a method of analyzing multiple models of a disease or condition of interest is used, and the method correlates the results obtained from these different models; this minimizes possible errors that might be introduced by choosing a model for a particular disease . This approach minimizes the impact of reasonable errors in the estimates of prevalence, allele frequencies, and ORs derived from sources on the calculation of relative risk. Due to the "linear" or monotonic nature of the effect of prevalence estimates on RR, incorrectly estimating prevalence has little or no effect on the final score; the same model is assumed to be applied consistently to all individuals generating reports.

In another embodiment, a method that considers environmental/behavioral/population data as additional "loci" is used. In related embodiments, these data may be derived from information sources such as medical or scientific literature or databases (eg, association of smoking w/lung cancer or from insurance industry health risk assessments). In one embodiment, a GCI score is generated for one or more complex diseases. Complex diseases can be influenced by multiple genes, environmental factors and their interactions. When studying complex diseases, a large number of possible interactions need to be analyzed. In one embodiment, a procedure such as Bonferroni correction is used to correct for multiple comparisons. In an alternative embodiment, the Simes test is used to control the overall significance level (also called "family error rate") when the tests are independent or show a particular type of dependence (Sarkar S. (1998)). Some probability inequalities for ordered MTP2 random variables: A proof of the Simes hypothesis (AnnStat26:494-504). If for any k in 1,...,K, p(k) ≤ αk/K, then the Simes test rejects the global null hypothesis for all K-test specificity null hypotheses true (SimesRJ (1986) An improved Bonferroniprocedureformultipletestsofsignificance.Biometrika73:751 -754).

Other embodiments, which can be used in the context of multigene and multienvironmental factor analyses, control the false-discovery rate, the expected proportion of falsely rejected null hypotheses. This approach is particularly beneficial when, as in microarray studies, part of the null hypothesis can be assumed to be wrong. Devlin et al. (2003, Analysis of multilocus models of fassociation. Genet Epidemiol 25:36-47) proposed Benjamini and Hochberg (1995, Controlling the false discovery rate: practical and powerful approach to multiple testing. JRStatSocSerB57 :289-300) A variant of the increment procedure. The Benjamini and Hochberg procedure is related to the Simes test; setting k ^* =maxk such that p(k)≦αk/K rejects all k ^* null hypotheses corresponding to p(1), . . . , p(k ^* ). In fact, the Benjamini and Hochberg procedure reduces to a Simes test when all null hypotheses are true (Benjamini Y, Yekutieli D (2001) The control of the false discovery rate in multiple testing under dependency. Ann Stat 29: 1165-1188).

In some embodiments, individuals are ranked based on their median score compared to a population of individuals to produce a final score, which can be expressed as a rank within the population, such as the 99th percentile or 99th, 98th, 97th, 96th, 95, 94, 93, 92, 91, 90, 89, 88, 87, 86, 85, 84, 83, 82, 81, 80, 79, 78, 77, 76, 75, 74, 73, 72, 71, 70, 69, 65, 60, 55, 50, 45, 40, 40, 35, 30, 25, 20, 15, 10, 5, or 0 percentiles. In another embodiment, the scores may be displayed as a range, such as the 100th to 95th percentile, the 95th to 85th percentile, the 85th to 60th percentile, or any subrange between the 100th to 0 percentile. In yet another embodiment, individuals are ranked by quartile, such as the highest 75th quartile or the lowest 25th quartile. In further embodiments, individuals are ranked compared to the mean or median score in the population.

In one embodiment, the population to which individuals are compared includes a large number of people from different geographic and ethnic backgrounds, such as a global population. In other embodiments, the population to which individuals are compared is limited to a particular geography, pedigree, race, sex, age (fetal, neonatal, child, juvenile, adolescent, adult, elderly individual), disease state (e.g., symptomatic of, asymptomatic, carrier, early onset, late onset). In some embodiments, the population to which the individual is compared is derived from information reported from public and/or private sources.

In one embodiment, the individual's GCI score or GCIPlus score is visualized using a display device. In some embodiments, a display screen (eg, a computer monitor or television screen) is used for visual display, such as a personal portal with relevant information. In another embodiment, the display device is a static display device, such as a printed page. In one embodiment, a display may include, but is not limited to, one or more of the following: bins (eg, 1-5, 6-10, 11-15, 16-20, 21-25, 26 -30, 31-35, 36-40, 41-45, 46-50, 51-55, 56-60, 61-65, 66-70, 71-75, 76-80, 81-85, 86-90 , 91-95, 96-100), color or grayscale gradients, thermometers, scales, pie charts, histograms or bar charts. For example, Figures 18 and 19 are different displays for MS and Figure 20 is for Crohn's disease. In another embodiment, a thermometer is used to display GCI score and disease/condition prevalence. In another embodiment, a thermometer shows levels as a function of reported GCI score, eg, Figures 15-17, with colors corresponding to risk. The thermometer can show a change in color with increasing GCI score (eg, a gradual change from blue at lower GCI scores to red at higher GCI scores). In a related embodiment, the thermometer displays level as a function of reported GCI score and color change as risk level increases.

In an alternative embodiment, the individual's GCI score is communicated to the individual using auditory feedback. In one embodiment, the auditory feedback is a verbal indication of whether the risk level is high or low. In another embodiment, the auditory feedback is a narrative of a particular GCI score, such as a number, percentile, range, quartile, or comparison to a population mean or median GCI score. In one embodiment, the living person delivers the auditory feedback in person or through a communication device, such as a telephone (landline, cellular or satellite phone), or through a personal portal. In another embodiment, auditory feedback is delivered by an automated system such as a computer. In one embodiment, auditory feedback is delivered as part of an Interactive Voice Response (IVR) system, a technology that allows a computer to detect speech and touchtones using normal phone calls. In another embodiment, individuals can interact with the central server through an IVR system. IVR systems can respond to pre-recorded or dynamically generated audio to interact with individuals and provide them with auditory feedback of their risk level. In one embodiment, the individual can call a number answered by the IVR. After optionally entering an authentication code, a security code, or going through a speech recognition process, the IVR system lets the subject select an option from a menu, such as a touch tone or spoken menu. One of these options may provide the individual with his or her risk level.

In another embodiment, an individual's GCI score is visualized using a display device and delivered using auditory feedback, such as through a personal portal. This combination may include a visual display of the GCI score and auditory feedback that discusses the relevance of the GCI score to the individual's overall health and possible preventive measures that may be suggested.

In one embodiment, a multi-step approach is used to generate the GCI score. Initially, for each state to be studied, the relative risk from the odds ratio for each genetic marker is calculated. For each prevalence value of p = 0.01, 0.02, ..., 0.5, the GCI score for the HapMap CEU population was calculated based on prevalence and HapMap allele frequencies. If the GCI score is constant under varying prevalence, the only assumption considered is the existence of a constructive model. Additionally, it can be determined that the model is sensitive to popularity. For any combination of uncalled values, the distribution of relative risk and score across the HapMap population is obtained. For each new individual, the individual score is compared to the HapMap distribution and the resulting score is the individual's rank in this population. Reported scores may be of lower resolution due to assumptions made during the process. The population will be divided into loci (3-6 bins), and the reported bin will be the one into which the individual rank falls. The number of bins may be different for different diseases based on considerations such as the resolution of the scores for each disease. In case of linkage between scores of different HapMap individuals, the average rank will be used.

In one embodiment, a higher GCI score is interpreted as indicating an increased risk of acquiring or being diagnosed with a condition or disease. In another embodiment, a mathematical model is used to derive the GCI score. In some embodiments, the GCI score is based on a mathematical model that accounts for the incomplete characteristics that underlie information about a population and/or a disease or condition. In some embodiments, the mathematical model includes certain at least one assumption as part of the basis for calculating the GCI score, wherein the assumptions include, but are not limited to: an assumption that an odds ratio is given; an assumption that the prevalence of a state is known; The assumption that the genotype frequencies in the population are known; and the assumption that the consumer comes from the population used in the study and the same family background as the HapMap; the assumption that the pooled risk is the product of the different risk factors for individual genetic markers. In some embodiments, a GCI can also include the assumption that the polygenotype frequency of a genotype is the product of the allele frequencies of each SNP or individual genetic marker (eg, different SNPs or genetic markers are independent across the population).

Productive model

In one embodiment, the GCI score is calculated under the assumption that the risk attributable to a set of genetic markers is the product of the risks attributable to individual genetic markers. This means that different genetic markers are attributable to the risk of disease independently of other genetic markers. Formally, there are _k genetic markers with risk alleles r ₁ , . . . , _rk and non-risk alleles n ₁ , . . . , nk. In SNPi we denote three possible genotype values r _i r _i , _ni r _i and n _{i ni} . The genotype information of an individual can be described by a vector (g ₁ , . . . , g _k ), where g _i can be 0, 1 or 2 according to the number of risk alleles at position i. we pass Indicates the relative risk of a heterozygous genotype at the same position compared to a homozygous non-risk allele at position i. In other words, we define Similarly, we denote the relative risk of r _i r _i genotype as Under the productive model, we assume that the risk of an individual with a genotype (g ₁ ,...,g _k ) is Active models have been used previously in the literature to simulate case-control studies or for visualization purposes.

Assess relative risk

In another embodiment, the relative risk for different genetic markers is known, and a cumulative model can be used for risk assessment. However, in some embodiments including association studies, the study design prevents reporting of relative risk. In some case-control studies, relative risks could not be calculated directly from the data without further assumptions. Instead of reporting relative risk, the usual way is to report the odds ratio (OR) of the genotype, which is the probability of carrying a given risk genotype disease (r _i r _i or _ni r _i ) versus not carrying a given risk genotype disease The ratio of the probability of . formal,

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}$

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}$

Finding relative risk from odds ratios may require additional assumptions. For example, assuming the allele frequencies in the entire population and known or evaluated (these can be evaluated from existing datasets such as the HapMap dataset including 120 chromosomes), and/or assume that the prevalence p=p(D) of the disease is known. From the above three equations, we can get:

p=a·P(D|n _i n _i )+b·P(D|n _i r _i )+c·P(D|r _i r _i )

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}$

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; )</span>}$

Through the definition of relative risk, after dividing by the pP(D|n _{i ni} ) term, the first equation can be rewritten as:

$\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; ,</span>$

And thus, the latter two equations can be rewritten as:

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\frac{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}}$

(1)

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\frac{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}}$

It should be noted that when a=1 (the non-risk allele frequency is 1), the equation system 1 is equivalent to the Zhang and Yu formula in ZhangJ and YuK. 1, 1998, the entire contents of which are incorporated by reference). Contrary to the Zhang and Yu formula, some embodiments of the invention take into account allele frequencies in a population, which may affect relative risk. Still other embodiments take into account the interdependence of relative risks. This is in contrast to calculating each relative risk independently.

System of equations 1 can be rewritten as two quadratic equations with up to four possible solutions. A gradient descent algorithm (gradient descent algorithm) can be used to solve these equations, where the starting point is set to the odds ratio, e.g., and .

For example:

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

Finding solutions to these equations amounts to finding the minimum of the function g(λ ₁ , λ ₂ )=f ₁ (λ ₁ ,λ ₂ ) ² +f ₂ (λ ₁ ,λ ₂ ) ² .

therefore,

$\frac{\mathrm{<span\; class="notranslate">\; dg</span>}}{{\mathrm{<span\; class="notranslate">\; d\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>$

$\frac{\mathrm{<span\; class="notranslate">\; dg</span>}}{{\mathrm{<span\; class="notranslate">\; d\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; c\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; b\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; c</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>$

In this example, we start by setting x ₀ =OR ₁ , y ₀ =OR ₂ . We set the value [ε]= ^10-10 as the tolerance constant of the whole algorithm. In iteration i, we define $\mathrm{\&gamma;}=\min \{0.001,\frac{x_{i-1}}{\left[\mathrm{epsilon}\right]+10\left|\frac{\mathrm{d\; g}}{{\mathrm{d\&lambda;}}_1}(x_{i-1},{\mathrm{the\; y}}_{i-1})\right|},\frac{{\mathrm{the\; y}}_{i-1}}{\left[\mathrm{epsilon}\right]+10\left|\frac{\mathrm{d\; g}}{{\mathrm{d\&lambda;}}_2}(x_{i-1},{\mathrm{the\; y}}_{i-1})\right|}\}.$ Then, we set

${\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\frac{\mathrm{<span\; class="notranslate">\; dg</span>}}{{\mathrm{<span\; class="notranslate">\; d\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&gamma;</span>}\frac{\mathrm{<span\; class="notranslate">\; dg</span>}}{{\mathrm{<span\; class="notranslate">\; d\&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>$

These iterations are repeated until g(x _i , y _i ) < tolerance, where the tolerance is set to 10 ⁻⁷ in the code provided.

In this example, these equations give positive solutions for different values of a, b, c, p, OR ₁ and OR ₂ . Figure 10

Robustness of relative risk assessment

In some embodiments, the effect of different parameters (prevalence, allele frequency, and odds ratio error) on the estimate of relative risk is determined. To determine the effect of allele frequency and prevalence estimates on relative risk values, the relative risk (under HWE) from a set of values for different odds ratios and different allele frequencies was calculated, and the results of these calculations for Plotting popularity values ranging from 1 to 1. Figure 10. Alternatively, for a fixed prevalence value, the resulting relative risk can be plotted as a function of risk allele frequency. Figure 11. When p=0, λ ₁ =OR ₁ , and λ ₂ =OR ₂ , and when p=1, λ ₁ =λ ₂ =0. This can be calculated directly from the equation. Additionally, in some embodiments, when the risk allele frequency is high, λ ₁ is closer to a linear function, and λ ₂ is closer to a concave function with a bounded second derivative. In the limit, when c=1, λ ₂ =OR ₂ +p(1-OR ₂ ), and If OR ₁ ≈OR ₂ , the latter is also close to a linear function. When the risk allele frequency is low, λ1 and _λ2 approximate the behavior of the function ₁ /p. In the limit case, when c=0, ${\mathrm{\&lambda;}}_1=\frac{o{R}_1}{1-p+\mathrm{pO}{R}_1},$ ${\mathrm{\&lambda;}}_2=\frac{o{R}_2}{1-p+\mathrm{pO}{R}_2}.$ This suggests that for high risk allele frequencies, incorrect prevalence estimates will not significantly affect the resulting relative risk. Alternatively, for low risk allele frequencies, if the correct prevalence p is replaced by the prevalence value p' = αp, the resulting relative risk is eliminated by at most coefficient. This is illustrated in panels (c) and (d) of FIG. 11 . It should be noted that for high risk allele frequencies the two plots are quite similar, while for low allele frequencies there is a higher bias in the difference in relative risk values, which is less than a factor of two.

Calculate GCI Score

In one embodiment, a genetic composite index is calculated using a reference set representative of a relevant population. This reference set can be one of the populations in HapMap or another genotype dataset.

In this embodiment, the GCI is calculated as follows. For each of the k risk loci, the relative risk was calculated from the odds ratio using Equation System 1. Then, calculate the product score for each individual in the reference set. The GCI of an individual with a productive score s is the score of all individuals in the reference dataset with a score of s' ≤ s. For example, if 50% of individuals in the reference set have a product score less than s, then the final GCI score for that individual will be 0.5.

other models

In one embodiment, a product model is used. In alternative embodiments, other models may be used for the purpose of determining the GCI score. Other suitable models include, but are not limited to:

Additive model. Under an additive model, the risk hypothesis for an individual with a genotype (g ₁ , ... g _k ) is $\mathrm{GCI}(g_1,...,g_k)=\underset{i=1}{\overset{k}{\mathrm{\&Sigma;}}}f\left({\mathrm{\&lambda;}}_{g_i}^i\right).$

Generalized additive models. In a generalized additive model, it is assumed that there exists a function f such that the risk of an individual with a genotype (g ₁ ,...g _k ) is

Harvard modified score (Het). This score was derived by G. AColditz et al. and thus applied to genetic markers (Harvard report on cancer prevention volume 4: Harvard cancer risk index. Cancer Causes and Controls, 11:477-488, 2000, the entire content of which is hereby incorporated). Although the function f operates on odds ratios rather than relative risks, the Het score is essentially a generalized additive score. This is useful in situations where relative risk is difficult to assess. To define a function f, an intermediate function g is defined as:

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}\mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 1.09</span>}\\ \mathrm{<span\; class="notranslate">\; 5</span>}& \mathrm{<span\; class="notranslate">\; 1.09</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 1.49</span>}\\ \mathrm{<span\; class="notranslate">\; 10</span>}& \mathrm{<span\; class="notranslate">\; 1.49</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 2.99</span>}\\ \mathrm{<span\; class="notranslate">\; 25</span>}& \mathrm{<span\; class="notranslate">\; 2.99</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 6.99</span>}\\ \mathrm{<span\; class="notranslate">\; 50</span>}& \mathrm{<span\; class="notranslate">\; 6.99</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; x</span>}\end{array}\right.$

then calculate amount of which is the frequency of SNPi heterozygous individuals in the entire reference population. The function f is then defined as f(x)=g(x)/het, and the Harvard modified score (Het) is simply defined as

Harvard modified score (Hom). In addition to the value het is valued This score is similar to the Het score except instead of , where, is the frequency of individuals homozygous for the risk allele.

maximum odds ratio. In this model, it is assumed that one of the genetic markers (the one with the largest odds ratio) gives a lower bound on the combined risk for the entire subject group. Formally, the score of an individual with genotype (g ₁ ,...g _k ) is

Comparison Between Ratings

In one embodiment, for the 10 T2D-associated SNPs, GCI scores were calculated based on multiple models on the entire HapMap CEU population. The associated SNPs are rs7754840, rs4506565, rs7756992, rs10811661, rs12804210, rs8050136, rs1111875, rs4402960, rs5215, rs1801282. For each of these SNPs, the odds ratios of the three possible genotypes are reported in the literature. The CEU cohort consisted of thirty mother-father-child trios. To avoid dependencies, sixty parents from this group were employed. Excluding one individual with a no call in one of the 10 SNPs resulted in a set of 59 individuals. GCI ranks for individual individuals were then calculated using several different models.

It can be observed that for this dataset, the different models produce highly correlated results. Figures 12 and 13. Spearman correlations were calculated between pairs of models (Table 2), which showed that the additive and additive models had a correlation coefficient of 0.97, and thus the GCI scores were robust when using the additive or additive models. Similarly, the correlation between the Harvard modified score and the integrative model was 0.83, and the correlation coefficient between the Harvard score and the additive model was 0.7. However, using the maximum odds ratio as a genetic score yields a dichotomous score defined by one SNP. Collectively, these results suggest that scoring rankings provide a stable framework that minimizes model dependencies.

Table 2: Spearman correlations of score distributions for CEU data between model pairs.

The effect of variation in T2D prevalence on the resulting distributions was determined. The prevalence values varied between 0.001 and 0.512 (Figure 14). For the T2D case, it can be seen that different prevalence values lead to the same order of individuals (Spearman correlation > 0.99), so an artificial fixed value of 0.001 for popularity can be assumed.

Extend the model to any number of variants

In another embodiment, the model can be extended to situations where any number of possible variants occur. The previous considerations dealt with the case where there are three possible variants (nn, nr, rr). In general, any number of variants can be found in a population when multiple SNP associations are known. For example, when an interaction between two genetic markers is associated with a state, there are nine possible variants. This resulted in eight different odds ratio values.

To generalize the original formula, it can be assumed that there are k+1 possible variants a ₀ , ..., a _k , with frequencies f ₀ , f ₁ , ..., f _k , with an odds ratio of 1 determined, OR ₁ , ..., OR _k and unknown relative risk values 1, λ ₁ , ..., λ _k . It can be further assumed that all relative risks and odds ratios are determined relative to a ₀ , and therefore, ${\mathrm{\&lambda;}}_i=\frac{P\left(\mathrm{D.}\right|a_i)}{P\left(\mathrm{D.}\right|a_o)}$ and ${\mathrm{OR}}_i=\frac{P\left(\mathrm{D.}\right|a_i)}{P\left(\mathrm{D.}\right|a_o)}\&Center\; Dot;\frac{1-P\left(\mathrm{D.}\right|a_i)}{1-P\left(\mathrm{D.}\right|a_o)}.$ based on:

$\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$

can be sure

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; .</span>$

Also, if you set This leads to the following equation:

${\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; ,</span>$

and therefore,

$\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; ,</span>$

or

$\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; p</span>}}<span\; class="notranslate">\; .</span>$

The latter is an equation with one variable (C). This equation can yield many different solutions (essentially, up to k+1 different solutions). Standard optimization tools (such as gradient descent) can be used to find the closest solution to C ₀ =∑f _i t _i .

The present invention uses a robust scoring framework for risk factor quantification. Although different genetic models can lead to different scores, the results are generally correlated. Therefore, the quantification of risk factors is generally independent of the model used.

Evaluating relative risk case-control studies

Also provided in the present invention is a method for estimating relative risk from multi-allelic odds ratios in case-control studies. In contrast to previous approaches, this approach takes into account allele frequencies, disease prevalence and dependencies among the relative risks of different alleles. The performance of the method was measured for a simulated case-control study and found to be extremely accurate.

method

In the case of testing the association of a particular SNP with disease D, R and N represent the risk and non-risk alleles of this particular SNP. P(RR|D), P(RN|D) and P(NN|D) denote the assumption that the individual is homozygous for the risk allele, heterozygous or homozygous for the non-risk allele, respectively Probability of being affected by disease. f _RR , f _RN and f _NN are used to represent the frequencies of the three genotypes in the population. Using these definitions, relative risk is defined as

${\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; RR</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span>}$

${\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; RN</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span>}$

In case-control studies, P(RR|D), P(RR|~D) values (ie, the frequency of RRs in cases and controls), and P(RN|D), P(RN|~D), P(RN|~D), P(NN|D) and P(NN|～D), that is, the frequency of RN and NN in cases and controls. To estimate relative risk, Bayes' law can be used to obtain:

${\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; RR</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}}$

${\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; RN</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}}$

Thus, if the frequencies of genotypes are known, one can use them to calculate relative risk. The frequencies of genotypes in a population cannot be calculated from case-control studies themselves because they depend on the prevalence of the disease in the population. In particular, if the prevalence of a disease is p(D), then:

f _RR =P(RR|D)p(D)+P(RR|～D)(1-p(D))

f _RN =P(RN|D)p(D)+P(RN|～D)(1-p(D))

f _NN =P(NN|D)p(D)+P(NN|~D)(1-p(D)).

When p(D) is small enough, the frequency of the genotype can approach that in the control population, but when the prevalence is high, this will not be an accurate estimate. However, given a reference dataset (eg, HapMap[cite]), one can estimate genotype frequencies based on the reference dataset.

Most recent studies do not use reference data sets to estimate relative risks and report only odds ratios. odds ratio can be written

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; RR</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; ~</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; RR</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; ~</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}$

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; RN</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; ~</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; RN</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; ~</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}$

Odds ratios are often favorable since it is generally not necessary to have estimates of allele frequencies in a population; to calculate odds ratios, all that is usually required are the genotype frequencies in cases and controls.

In some cases, genotype data per se are not available, but summary data (eg odds ratios) are available. This is the case when a meta-analysis is performed based on results from previous case-control studies. In this case, it was demonstrated how to find the relative risk from the odds ratio. The facts shown using the following equation:

p(D)=f _RR P(D|RR)+f _RN P(D|RN)+f _NN P(D|NN)

If this equation is divided by P(D|NN), we get

$\frac{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}$

This allows the odds ratio to be written as:

$\begin{array}{c}{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; RR</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; RR</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}\frac{\frac{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\frac{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; NN</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}}<span\; class="notranslate">\; =</span>\\ {\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}}\end{array}$

Through similar calculations, the following equation system is obtained:

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}}$

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RR</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; NN</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; RN</span>}}}<span\; class="notranslate">\; .</span>$

Equation 1

Relative risk can be obtained by solving this system of equations if the odds ratio, the genotype frequency in the population, and the prevalence of the disease are known.

Note that there are two quadratic equations, so they have a maximum of four solutions. However, as shown below, there is usually one possible solution to this equation.

It should be noted that when f _NN = 1, Equation System 1 is equivalent to the Zhang and Yu formula; however, the allele frequencies in the population are considered here. Furthermore, our method takes into account the fact that two relative risks are dependent on each other, whereas previous methods propose to calculate each relative risk independently.

Relative risk of multiallelic loci. Calculations are slightly more complicated if multiple markers or other multi-allelic variants are considered. a ₀ , a ₁ , . . . , a _k represent possible k+1 alleles, where a ₀ is a non-risk allele. Allele frequencies f ₀ , f ₁ , f ₂ , . . . , f _k in the population for k+1 possible alleles are assumed. For allele i, the relative risk and odds ratios are defined as

${\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span>}$

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\frac{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}$

The following equation applies to the prevalence of a disease:

$\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

Therefore, by dividing both sides of the equation by p(D|a ₀ ), we get:

$\frac{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}$

get:

${\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; ,</span>$

by setting $C=\underset{i=0}{\overset{k}{\mathrm{\&Sigma;}}}{f}_i{\mathrm{\&lambda;}}_i,$ get ${\mathrm{\&lambda;}}_i=C\&CenterDot;\frac{{\mathrm{OR}}_i}{p\left(\mathrm{D.}\right){\mathrm{OR}}_i+C-p\left(\mathrm{D.}\right)}.$ Therefore, by the definition of C, it follows that:

$\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\frac{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\frac{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; OR</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; D.</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; .</span>$

This is a polynomial equation with one variable C. Once C is determined, the relative risk is determined. The polynomial is of degree k+1, and thus we expect to have at most k+1 solutions. However, since the right-hand side of the equation strictly reduces to a function of C, there may usually only be one solution to this equation. This solution is easy to find using a binary search, since the solution is bounded by C = 1 and $C=\underset{i=0}{\overset{k}{\mathrm{\&Sigma;}}}o{R}_i$ between.

Stability of relative risk assessments. The effect of various parameters (prevalence, allele frequency, and odds ratio error) on estimates of relative risk was determined. To determine the effect of allele frequency and prevalence estimates on relative risk values, the relative risk was calculated from a set of values (under HWE) for different odds ratios, different allele frequencies, and for values ranging from 0 to 1 The popularity values plot the results of these calculations.

Additionally, the resulting relative risk is plotted as a function of risk allele frequency for a fixed prevalence value. It is clear that λ _RR =OR _RR and λ _RN =OR _RN in all cases of p(D)=0, and λ _RR =λ _RN =0 when p(D)=1. This can be directly calculated from Equation 1. In addition, λ _RR approaches a linear behavior when the risk allele frequency is high, and λ _RN approaches a concave function with bounded second derivative. When the risk allele frequency is low, λ _RR and λ _RN are close to the performance of function 1/p(D). This means that for high risk allele frequencies, erroneous estimates of prevalence will not greatly affect the resulting relative risk.

The following examples illustrate and explain the invention. The scope of the present invention is not limited to these examples.

Example I

Generation and analysis of SNP distribution maps

The individual is provided with a sample tube in a kit (eg, available from DNAGenotek) into which the individual deposits a saliva sample (approximately 4 ml) from which genomic DNA will be extracted. Saliva samples were delivered to a CLIA-certified laboratory for processing and analysis. Typically, the sample is delivered to the testing facility by overnight mail in a shipping container conveniently provided to the individual within the collection kit.

In a preferred embodiment, genomic DNA is isolated from saliva. For example, using the DNA self-collection kit technology provided by DNAGenotek, individuals collect approximately 4 ml of saliva samples for clinical processing. After delivery of the sample to an appropriate laboratory for processing, DNA is isolated by heat denaturation and protease digestion of the sample (usually at 50°C for at least one hour using reagents provided by the collection kit supplier). Subsequently, the samples were centrifuged and the supernatant was subjected to ethanol precipitation. The DNA pellet was suspended in a buffer suitable for subsequent analysis.

Genomic DNA of the individual is isolated from the saliva sample according to known procedures and/or as provided by the manufacturer of the collection kit. Typically, samples are first subjected to heat denaturation and protease digestion. Next, the samples were centrifuged and the supernatant was retained. The supernatant was then subjected to ethanol precipitation to obtain a precipitate containing approximately 5-16 ug of genomic DNA. The DNA pellet was suspended in 10 mM Tris (pH 7.6), 1 mM EDTA (TE). SNP profiles are generated by hybridizing genomic DNA to commercially available high-density SNP arrays, such as those provided by Affymetrix or Illumina, using the instrumentation and instructions provided by the array manufacturer. Store individual SNP profiles in an encrypted database or vault.

The patient data structure is queried for risk-conferring SNPs by comparison to clinical databases of established, medically relevant SNPs whose presence in the genome correlates with a given disease or condition. The database includes information on the statistical association of specific SNPs and SNP haplotypes with specific diseases or conditions. For example, as shown in Example III, polymorphisms in the apolipoprotein E gene result in distinct isoforms of the protein, which in turn correlates with the statistical likelihood of developing Alzheimer's disease. As another example, individuals with a variant of the coagulation protein Factor V called Factor V Leiden have an increased tendency to coagulate. A number of genes in which SNPs are associated with disease or condition phenotypes are shown in Table 1. The scientific accuracy and significance of the information in the database is approved by the Research/Clinical Advisory Board and may be checked by overseeing government agencies. The database can be continuously updated as more SNP-disease associations emerge from the scientific community.

The analysis results of individual SNP profiles are securely provided to patients via online portal or email. Provide explanations and supportive information to the patient, such as the information on Factor V Leiden shown in Example IV. Secure access to an individual's SNP profile information (eg, through an online portal) would facilitate discussions with the patient's physician and empower options for personalized medicine.

Example II

Updating of Genotype Correlations

In response to a request to initially determine an individual's genotype correlation, a genomic profile is generated, the genotype correlation is obtained, and the results are provided to the individual as described in Example 1. After an initial determination of an individual's genotype correlations, later when additional genotype correlations are known, updated correlations are determined or can be determined. Registered users have advanced registration and their genotype profiles are kept in an encrypted database. The updated correlations are performed on the stored genotype profiles.

For example, as described in Example 1 above, initial genotype correlations have determined that a particular individual does not have ApoE4, and therefore is not susceptible to early-onset Alzheimer's disease, and that this individual is determined not to have factor VLeiden. After this initial determination, new associations become known and validated such that a polymorphism in a given gene (let's say gene XYZ) is associated with a given state (let's say state 321). Add this new genotype correlation to the master database of human genotype correlations. Updates are then provided to a specific individual by first fetching the data for the associated gene XYZ from the specific individual's genomic profile stored in an encrypted database. The associated gene XYZ data for a particular individual is compared with the updated master database's gene XYZ information. From this comparison, a particular individual's susceptibility or predisposition to state 321 is determined. The results of this determination are added to the genotype correlations for a particular individual. An updated result of whether a particular individual is susceptible to status 321 or is genetically susceptible is provided to the particular individual along with explanatory and supportive information.

Example III

Association between the ApoE4 locus and Alzheimer's disease

Alzheimer's disease (AD) risk has been shown to be associated with a polymorphism in the apolipoprotein E (APOE) gene, which results in three isoforms of APOE called ApoE2, ApoE3 and ApoE4 . These isoforms differ from each other by one or two amino acids at residues 112 and 158 of the APOE protein. ApoE2 contains cysteine/cysteine at position 112/158; ApoE3 contains cysteine/arginine at position 112/158; and ApoE4 contains arginine/arginine at position 112/158. As shown in Table 3, the risk of Alzheimer's disease onset at a younger age increases with the copy number of the APOEε4 gene. Also, as shown in Table 3, the relative risk of AD increased with the copy number of APOEε4 gene.

Table 3: Prevalence of AD risk alleles (Corder et al., Science: 261:921-3, 1993)

Table 4: Relative risk of AD with ApoE4 (Farrer et al., JAMA: 278: 1349-56, 1997)

APOE genotype odds ratio ε2ε2 0.6 ε2ε3 0.6 ε3ε3 1.0 ε2ε4 2.6 ε3ε4 3.2 ε4ε4 14.9

Example IV

Information for Factor V Leiden positive patients

The following information is an example of information that might be provided to an individual with a genomic SNP profile showing the presence of the Factor VLeiden gene. The individual may have a base registration that may provide information in the initial report.

What is Factor VLeiden?

Factor VLeiden is not a disease, it refers to the presence of specific genes that are inherited by a person's parents. Factor V Leiden is a variant of the protein Factor V(5) required for blood clotting. People with a factor V deficiency are more likely to bleed severely, while those with factor V Leiden have an increased tendency for their blood to clot.

People who carry the factor V Leiden gene have a five-fold higher risk of developing blood clots (thrombosis) than the rest of the population. But many people with the gene never develop blood clots. In the UK and US, 5% of the population carry one or more Factor V Leiden genes, which is far more than the number of people who will actually develop thrombosis.

How do you get factor VLeiden?

The Factor V gene is inherited from a person's parents. As with all inherited traits, one gene is inherited from the mother and one from the father. Thus, it is possible to inherit: two normal genes or one factor VLeiden gene and one normal gene or two factor VLeiden genes. Having one gene for Factor VLeiden will result in a slightly higher risk of developing thrombosis, but having two genes will result in a much greater risk.

What are the symptoms of Factor V Leiden?

Have no symptoms unless you have a blood clot (thrombosis).

What are the red flags?

The most common problem is blood clots in the legs. Swelling, pain, and redness in the legs show this problem. In rarer cases, a blood clot in the lungs (pulmonary thrombosis) may develop, which causes difficulty breathing. Depending on the size of the clot, the severity of the condition can range from barely noticeable to severe breathing difficulties for the patient. In even rarer cases, a blood clot may occur in the arm or other body part. Because these clots form in the veins that carry blood to the heart rather than in the arteries (which pump blood from the heart), Factor VLeiden does not increase the risk of coronary thrombosis.

What can be done to avoid blood clots?

Factor VLeiden only slightly increases the risk of blood clots, and many people with this condition never develop blood clots. There are many things a person can do to avoid causing a blood clot. Avoid standing or sitting in the same position for long periods of time. When traveling long distances, it is important to exercise regularly - it is essential that the blood does not "stand still". Staying up late or smoking will greatly increase the risk of blood clots. Women who carry the factor VLeiden gene should not take birth control pills as this significantly increases the chance of developing blood clots. Women who carry the factor VLeiden gene should also consult their doctor before becoming pregnant as this also increases the risk of blood clots.

How do doctors find out if you have Factor VLeiden?

The gene for factor VLeiden can be found in blood samples.

Blood clots in the legs or arms are usually identified by ultrasound.

Blood clots can also be detected by X-rays after a substance is injected into the blood to make them visible. Blood clots in the lungs are more difficult to find, but usually doctors will use radioactive material to test the distribution of blood flow in the lungs and the distribution of air to the lungs. The two distribution patterns should match - a mismatch indicates the presence of a blood clot.

How is factor VIeiden handled?

People with factor VLeiden do not need treatment unless their blood begins to clot, in which case a doctor will prescribe a blood-thinning (anticoagulant) drug such as warfarin (for example, coumarin) Or heparin to prevent further blood clots. Treatment will usually last three to six months, but it may take longer if several blood clots are present. In severe cases, the course of drug therapy may continue indefinitely; in very rare cases, the blood clot may require surgical removal.

How is Factor VLeiden handled during pregnancy?

Women who carry both factor VLeiden genes will need to be treated with the clot-promoting drug heparin during pregnancy. The same treatment is given to women who carry only one gene for factor VLeiden who have had blood clots themselves or have a family history of blood clots.

All women who carry the factor VLeiden gene may need to wear special stockings during the second half of pregnancy to prevent blood clots. After the baby is born, they can be prescribed the anticoagulant drug heparin.

Prognosis

The risk of blood clots increases with age, but in an age-related survey of 100 people who carried the gene, only a minority were found to have ever had a blood clot. The National Society for Genetic Counselors (NSGC) can provide a list of genetic counselors in your area and information on establishing a family history. Search their online database at www.nsgc.org/consumer.

While preferred embodiments of the present invention have been shown and described herein, it will be clear to those skilled in the art that these embodiments are provided by way of example only. Numerous variations, changes and substitutions will now occur to those skilled in the art without departing from the invention. It should be understood that many alternatives to the embodiments of the invention described herein may be employed in practicing the invention. It is intended that the following claims define the scope of the invention and that the invention cover methods and structures within the scope of these claims and their equivalents.

Claims (20)

1. A method of generating a Genetic Composite Index (GCI) score for an individual, the method comprising: a) obtaining a genetic sample of said individual; b) generating a genome profile from said genetic sample; c) determining a GCI score for the individual's phenotype by comparing the individual's genomic profile to a current database of human genotype correlations, wherein human genotype correlations are the correlation of genetic variants to phenotypes, wherein said GCI score is obtained by identifying an odds ratio or relative risk for one or more genetic loci, and wherein said phenotype is selected from Alzheimer's disease (AD), colorectal cancer (CRC), osteoarthritis (OA), exfoliation glaucoma (XFG), obesity (BMIOB), Graves' disease (GD) , hemochromatosis (HEM), myocardial infarction (MI), multiple sclerosis (MS), psoriasis (PS), restless legs syndrome (RLS), celiac disease (CelD), prostate cancer (PC), lupus ( SLE), macular degeneration (AMD), rheumatoid arthritis (RA), breast cancer (BC), Crohn's disease (CD), type 2 diabetes (T2D), and combinations thereof; And wherein said genetic variation is selected from SNP: rs4420638 when said genotype is AD, rs6983267 when said genotype is CRC, rs4911178 when said genotype is OA, and rs2165241 when said genotype is XFG , when the genotype is BMIOB, it is rs9939609 or rs9291171, when the genotype is GD, it is rs3087243, DRB1*0301DQA1*0501, when the genotype is HEM, it is rs1800562 or rs129128, when the genotype is MI, it is rs1866389 , rs1333049 or rs6922269, when the genotype is MS, it is rs6897932, rs12722489 or DRB1*1501, when the genotype is PS, it is rs6859018, rs11209026 or HLAC*0602, when the genotype is RLS, it is rs6904723, rs2300478, rs1026732 Or rs9296249, when the genotype is CelD, it is rs6840978, rs11571315, rs2187668 or DQA1*0301DQB1*0302, when the genotype is PC, it is rs4242384, rs6983267, rs16901979, rs17765344 or rs4430796, when the genotype is 17rs15LE , rs10954213, rs2004640, DRB1*0301 or DRB1*1501, when the genotype is AMD, it is rs10737680, rs10490924, rs541862, rs2230199, rs1061170 or rs9332739, when the genotype is RA, it is rs6679677, rs112033DR0, rs112033DR06 、DRB1*0401或DRB1*0404，所述基因型为BC时是rs3803662、rs2981582、rs4700485、rs3817198、rs17468277、rs6721996或rs3803662，所述基因型为CD时是rs2066845、rs5743293、rs10883365、rs17234657、rs10210302、rs9858542 , rs11805303, rs1000113, rs17221417, rs2542151 or rs10761659, when the genotype is T2D, it is rs13266634, rs4506565, rs10012946, rs7756992, rs10811661, rs12288738, rs80501136, rs 875, rs4402960, rs5215 or rs1801282; and d) Report the GCI score. 2. The method of claim 1, wherein a third party obtains the genetic sample. 3. The method of claim 1, wherein said generating a genomic profile is performed by a third party. 4. The method of claim 1, wherein said reporting comprises transmitting said results over a network. 5. The method of claim 1, wherein the genomic profile is the entire genome of the individual. 6. The method of claim 1, wherein the method comprises determining the plurality of relative risks or odds ratios from 10 or more genotype correlations. 7. The method of claim 1, further comprising generating a GCIplus score. 8. The method of claim 1, wherein the genetic sample is from a biological sample selected from the group consisting of blood, hair, skin, saliva, semen, urine, fecal matter, sweat, and oral samples. 9. The method of claim 1, wherein the genotype correlation is a correlation of a single nucleotide polymorphism with a phenotype of a non-medical condition. 10. The method of claim 1, wherein the genomic profile is generated using a high-density DNA microarray, genomic DNA sequencing, or a PCR-based method. 11. The method of claim 1, wherein the results further comprise incorporating data selected from the group consisting of physical data, medical data, demographic data, exposure data, lifestyle data, behavioral data, race, pedigree, geography, gender, age, familial Characterization of the individual by history and predetermined phenotype. 12. The method of claim 1, wherein the genomic profile includes genetic markers of linkage disequilibrium for genetic variants associated with a phenotype. 13. The method of claim 1, wherein the GCI score is an estimated lifetime risk. 14. The method of claim 1, wherein the genomic profile comprises at least 100,000 genetic variants. 15. The method of claim 1, wherein the genomic profile comprises at least 400,000 genetic variants. 16. The method of claim 1, further comprising reporting information about said phenotype, wherein the information is selected from the group consisting of preventive strategies, health information, therapy, symptom awareness, early detection protocols, intervention protocols, and accurate Identification and classification. 17. The method of claim 11, wherein the physical data is selected from the group consisting of blood pressure, heart rate, glucose level, metabolite level, ion level, weight, height, cholesterol level, vitamin level, blood count, body mass index (BMI) , protein and transcript levels. 18. The method of claim 1, further comprising: f) updating said database with at least one human genotype correlation; and g) generating at least one additional relative risk or odds ratio of said phenotype by comparing said individual's genomic profile with said at least one human genotype correlation of step f); h) calculating at least one updated genetic composite index (GCI) from said at least one additional relative risk or odds ratio determined in step g); and i) reporting said results from step h) to said individual or said individual's healthcare manager. 19. The method of claim 1, wherein the reporting of the at least one GCI score comprises a network transmission. 20. The method of claim 19, wherein the reporting includes transmitting the at least one GCI score via an online portal.