Multifactorial Traits
Chapter Seven
Multifactorial
• Not all diseases are Mendelian
• Multifactorial = many factors
In Genetics:
• Multifactorial = both environment and
genetics (usually more than one gene)
• Polygenic = more than one gene
• Each gene separately follows Mendel’s
laws, but the trait overall does not
Complex Traits
• Disorder that is proven heritable, yet has
no clear mode of inheritance
– Doesn’t follow Mendel’s laws
• More than one gene - polygenic
• Interaction between genes - epistasis
• Interaction between genes and
environment - multifactorial
Why Common Complex Disorders
and Rare Mendelian Disorders?
• Evolution can act upon a single
detrimental gene
– negative selection
• Gene functions that are good for some
things, but can be harmful in excess
ex rational fear vs. anxiety disorders
• Normal alleles only predisposing
– other mutations/environment present
Genes vs. Environment
100
Percent Genetic Control
50
Eye Color Heart Disease Car Accident
Percent Environmental Control
Genes vs. Environment
100
50
Eye Color Heart Disease Car Accident
Causation
• There are only two factors that cause
any trait:
1. Genetics
2. Environment
• Nature vs. Nurture
• Interaction of two is multifactorial
Multifactorial Traits:
• Height
• Weight
• Eye color
– although basically no environmental
component is involved
• Skin color
• Heart Disease
• What else?
Quantitative Traits
• Quantitative = measurable traits
• Disease state is Qualitative
– Affected vs. unaffected
– Show trait vs. do not show trait
ex Autistic or normal?
• Multifactorial traits are often Quantitative
– Height ? – quantify in inches
– Weight ? – quantify in pounds
ex “speaking ability is stuck at 2 years of age”
Quantitative Traits
• Quantitative traits produce a continuum of
phenotypes:
# Individuals
Height (in)
Quantitative Traits
• Disease state may be beyond “threshold”
# Individuals
Speaking Ability
Autistic
Quantitative Traits
Rather than genes people often talk about:
• Quantitative Trait Loci (QTL) =
chromosomal regions that have been
associated with a complex trait
• If a QTL is correct then one of the genes
residing in this region should be directly
involved in causing trait
Remember – More than one gene!
therefore – more than one QTL too
How to identify QTL
1. Linkage
2. Animal Models
3. Association Studies
QTL Mapping
• Start with a complex trait of interest
• Phenotype a large group of individuals for
trait – quantitatively
• Genotype everyone
• Do people who share the trait also share
specific genomic regions (QTL) more often
than chance?
Wait a Second!?!
• What if the trait we are interested in is
NOT genetic at all?
• What would happen then?
• No matter how hard you studied, you
would never find the genes, because there
are none!
st
1 : Is Trait Genetic?
• Calculate Heritability
• Heritability (H2) = proportion of the trait
that is controlled by genetics
– H2 = 100 % – trait is fully genetic
– H2 = 0 % – trait is fully environmental
• Complex traits are somewhere in between
Heritability
• Different ways to calculate/estimate
• All based on this:
# of relatives that share the trait
# expected to share if trait was 100% genetic
• Expected is calculated based on amount of
genetic material shared between two
relatives
Heritability – Expected:
• Use “Correlation Coefficient”
• Fancy way of saying – how related are
these two people?
Relationship Percent Genes Shared
Monozygotic Twins 100%
Dizygotic Twins 50%
Siblings 50%
Parent and Child 50%
Aunt or Uncle 25%
Grandparent 25%
First Cousin 12.5%
Heritability
• Gather a group of relatives
– 100 sibling pairs
– 100 MZ twins
• Phenotype for trait:
ex 40 of 100 sib pairs share trait
– .40/.50 = 80 %
– Therefore, trait is 80 % controlled by genetics
– 20 % controlled by environment
Problems with Heritability?
What do you think is wrong, or lacking, with
H2 estimate from relative pairs?
1. Families share genetics AND
environment
2. More shared genetics usually correlates
to more shared environment too:
ex Twins usually share bedroom
Siblings raised in same house, unlike cousins
Separate Genes from
Environment
1. Adopted children
• Compare birth parents to adopted parents
• Sharing environment only, not genes
2. Twins
• Compare MZ vs. DZ twins
• Differences in twins is only difference
between amount of genes shared,
environment is constant
Concordance Rates
• Compare the percent of MZ twins that are
concordant for the trait vs. the percent of
DZ twins concordant
• If MZ twins are more concordant than
difference is attributed to genetics
ex Autism: 90% MZ vs. 2% DZ
Homosexuality: 52% MZ vs. 20% DZ
Depression: 46% MZ vs. 9% DZ
Problem with Concordance?
More shared genetics usually correlates to
more shared environment too
– MZ twins can be raised in more similar
environments than DZ twins
ex Tuberculosis was found more often in MZ
twins than DZ twins
– Obviously TB (caused by bacteria), is not
genetic at all!
What would be ideal?
• MZ twins, separated at birth, raised
completely apart
• Animal models, where one can separate
the genetics from the environment
Genetic Counselors
• What happens when parents ask you the
risk of having a child with a complex
disorder?
• Cannot calculate an exact statistical
likelihood based on Mendel’s laws
Instead give:
• Heritability estimates
• Empiric Risk
Empiric Risk
• The incidence of the trait in a specific
population:
Differs based on ethnic background:
Ethnic Background Prevalence of Disease
Caucasian 50
African 3.8
African-American 26
Native American 5.9
Also depending on whether relative has trait:
Relationship to Affected Individual Risk of Recurrence
Identical Twin 40%
Sibling 5%
Child 3.5%
First Cousin 0.30%
QTL Mapping
• Start with a complex trait of interest
– That shows strong genetic contribution
• Phenotype a large group of individuals for
trait – quantitatively
• Genotype everyone
• Do people who share the trait also share
specific genomic regions (QTL) more often
than chance?
QTL Mapping
1. Linkage:
- Marker and phenotype are correlated
2. Animal Models
- Animal can be manipulated to show phenotype
- Compare differences in genetics, matched to
differences in phenotypes
3. Association Studies:
- Are certain alleles correlated with phenotype?
Association Studies
“Linkage Disequilibrium” – alleles are
inherited together (rather than genes)
– LD only ranges a short distance
– ~ 10,000 bases
– Because alleles are so close they are always
inherited together (no crossing over)
• Association comparing alleles
• Linkage usually done in families,
association usually done case vs. control
Case vs. Control
• Gather a group of individuals with trait
• Compare them to a group of individuals
who do not have trait
• Whatever alleles (genes) are different =
cause of trait
• Problems?
– What about ethnic genetic background?
– Need to perfectly match the background of
the two groups
Case vs. Control
• Perfect match = family members
Association studies are now often done in:
– Sib pairs (two siblings, both affected)
– Trios (two parents and one affected child)
• 2nd best = genotype both groups across
entire genome
– Make sure genetic backgrounds are the same
– “Genomic Controls”
Complexity of Trait
• More complex the trait, more difficult it will
be to identify QTL or genes involved
# Individuals Needed
1000
This is assuming
100 environment is
completely
50 controlled for
1 2 3 4 5
# Genes Causing Trait
Summary
1. Figure out whether trait is genetic:
• Compare relatives vs. general population
• MZ vs. DZ twins
• Adoption Studies
2. Estimate how genetic:
• Heritability
• Empiric Risk
3. Try to identify genes involved in trait
(QTL’s)
Summary
QTL = regions associated with phenotype:
– Linkage – correlation between genomic
regions and phenotype
– Association – correlation between alleles and
phenotype
– Animal Models
• Association Studies can be:
– Case vs. Control
– Sib pairs or Trios
Important Terms:
Empiric Risk – known incidence of trait in
particular population
Heritability – estimate of percentage of trait
that is caused by genetics
Correlation Coefficient – proportion of genes
two relatives share
Concordance – Percentage of twin pairs that
both show phenotype
Example: Obesity
• Qualitative = obese, overweight, normal
• Quantitative = weight in pounds
• Multifactorial
– Controlled by environment
– Controlled by genetics:
Next Class:
• Read Chapter Eight
• Homework – Chapter Seven Problems;
– Review: 1,4, 7,8,10
– Applied: 1,3,5,6,10,11
