EA047854B1 - METHODS OF NON-INVASIVE PRENATAL TESTING OF FETAL ANOMALIES - Google Patents

Abstract

The present invention relates to a method for detecting genetic and/or genomic abnormalities in a mixed sample, comprising the steps of biochemical and in silico enrichment of a subset of extracellular DNA fragments derived from the mixed sample. The invention utilizes a pool of long DNA probes to enrich sequences of interest in the mixed sample, followed by massively parallel sequencing and computer-based analysis of the enriched subset to identify the risk of genetic and/or genomic abnormalities in said subset of the mixed sample. The computer-based portion of the method does not require mandatory alignment to a reference genome or calibration values using reference samples. The method also includes a kit for implementing the invention.

Description

Field of technology

The present invention relates to the field of biology, medicine and chemistry, in particular to the field of molecular biology and more specifically to the field of molecular diagnostics.

State of the art

The detection of cell-free fetal DNA (cffDNA) in maternal plasma has greatly facilitated the development of non-invasive prenatal testing. However, most of the tests developed for the detection of fetal aneuploidies and microdeletions are based on single normalized values derived from read depth information. Although these tests can be considered as a significant improvement over existing methods, their clinical sensitivity does not exceed 99%. Especially when the proportion of cffDNA in the maternal bloodstream is below 4%, even with next-generation sequencing (NGS) technology, which has high sensitivity, achieving sufficient accuracy for non-invasive prenatal testing (NIPT) is challenging.

Brief description of the invention

According to the present invention, a method is provided for double enrichment of cell-free DNA (cfDNA) fragments originating from the placenta in a mixed biological sample containing cell-free fetal and cell-free maternal DNA using biochemical and in silico approaches that increase the signal-to-noise ratio by using long capture probes, fragment size analysis, hot spots of non-random fragmentation (HSNRF) and a new concept in bioinformatics. The method provides highly sensitive non-invasive detection of fetal abnormalities by using information obtained from enrichment of cell-free fetal DNA fragments in a prenatal sample using long DNA probes that capture cell-free DNA fragments and/or enrich for HSNRF.

The present invention also provides a novel computer-based method that organizes observed data (sequencing reads) into meaningful patterns that improve the signal-to-noise ratio when analyzing cell-free DNA samples containing a mixture of cfDNA fragments. In the method of the present invention, a plurality of DNA sequences are grouped in such a way that the degree of homology between said DNA sequences is maximized if they actually originate from the same structure, and minimized otherwise. Specific sequence patterns are identified in said data and used to distribute them into predetermined regions of interest. Said method can be used to detect specific patterns in DNA sequence data without the need for any prior knowledge, such as alignment to a human reference genome and/or calibration values using reference samples, and said method is used herein for non-invasive prenatal detection of chromosomal abnormalities in the fetus, such as aneuploidies, microdeletions, microduplications, and point mutations.

Thus, in a first aspect, the present invention relates to a method for detecting genetic/genomic abnormalities in a fetus in a mixed sample containing maternal and fetal cfDNA, said method comprising the following steps:

(i) obtaining a mixture of cfDNA from an individual;

(ii) preparation of a sequencing library from cfDNA;

(iii) enrichment of the cfDNA library using long DNA probes;

(iv) sequencing of the enriched cfDNA library;

(v) performing statistical analysis to determine the risk of chromosomal and/or other genetic abnormalities in fetal DNA;

wherein the risk of chromosomal and/or other genetic abnormality in fetal DNA is classified based on a double enrichment method comprising genomic hybridization and a computer-based method that increases the signal-to-noise ratio in said assay.

The method in which the enriched fragments are subjected to the procedure includes the steps of:

(i) grouping sequenced reads that can be paired based on nucleotide patterns, i.e., but not limited to, the degree of overlap in terms of nucleotide composition;

(ii) grouping and/or annotating at least a subset of the plurality of sequenced reads based on nucleotide patterns, i.e., but not limited to, sequence similarity of at least the 20 outermost nucleotides of the sequence;

(iii) matching the given nucleotide sequences with the reads according to (i) and (ii);

(iv) removal of duplicate sequenced reads;

(v) using the information obtained in steps (i) to (iv) to perform a statistical analysis to determine the risk of chromosomal and/or other genetic abnormalities in the fetal DNA.

In another aspect, the present invention provides a kit for implementing the invention. In one embodiment, the kit comprises:

a) probes that hybridize to at least one site in nucleic acid fragments

- 1 047 854 acids, wherein at least one of the said regions partially or completely covers a nucleic acid fragment where a hot spot of non-random fragmentation (HSNRF) is located, and, optionally,

b) reagents and/or software for implementing the invention and detecting genetic and/or genomic abnormalities in a sample.

Brief description of the figures

Figure 1 shows a proof-of-concept experiment in which 92 diploid samples and four trisomy 21 samples were classified using this method.

Figure 2 shows a proof-of-concept experiment in which 91 normal samples and two cases of chromosome 22q11.2 deletion syndrome were classified for microdeletion (chromosome 22q11.2 deletion syndrome) using this method.

Fig. 3 shows a kernel density function of the fragment sizes from a representative sample using (A) all fragments and (B) only fragments selected in this manner.

Detailed description of the invention

According to a first aspect of the present invention, the present application discloses a method for detecting a chromosomal abnormality in a mixed sample, comprising the steps of:

(a) obtaining a biological sample that contains a mixture of cell-free DNA (cfDNA) fragments, (b) preparing a sequencing library from the cfDNA fragments, (c) hybridizing one or more probes to at least one or more cfDNA fragments, (d) isolating from the library cfDNA fragments that bind to the probes, (e) sequencing the cfDNA fragments from the library that bind to the probes, (f) using information about the size, start and/or end of each or a subset of enriched cfDNA fragments according to steps (c-e) to select a fraction of cfDNA fragments hybridized to the one or more probes.

where step (f) involves calculating several statistical tests.

Thus, the present invention relates to a method for detecting genetic/genomic abnormalities in a fetus in a mixed sample containing maternal and fetal cfDNA, the method comprising the following steps:

(i) obtaining a mixture of cfDNA from an individual; (ii) preparing a sequencing library from the cfDNA; (iii) enriching the cfDNA library using long DNA probes; (iv) sequencing the enriched cfDNA library;

(v) performing statistical analysis to determine the risk of chromosomal and/or other genetic abnormalities in fetal DNA.

In the present application, HSNRF is defined as a genomic region containing, within a distance of less than 300 base pairs (bp) (preferably less than 200 bp, more preferably less than 100 bp), preferential sites that differentiate two tissue types present in a cfDNA mixture, and wherein said preferential sites are present in HSNRF regions with a higher frequency than in other non-HSNRF regions. In the present application, preferential sites are referred to as genomic bases in which the frequency of the read endpoint is significantly different (p-value of at least less than 0.05) between the two tissue types present in the cfDNA mixture.

In the present application, the mixture of nucleic acid fragments is preferably isolated from a sample taken from a eukaryotic organism, preferably a primate, more preferably a human.

In the context of the present invention, the expressions nucleic acid fragments and fragmented nucleic acids can be used interchangeably.

In one embodiment, the probes are long DNA molecules and:

(i) the length of each probe is from 100 to 500 base pairs, (ii) each denatured probe has a 5' end and a 3' end, (iii) preferably each probe binds to HSNRF at a distance of at least 10 base pairs at both the 5' end and the 3' end from regions bearing copy number variations (CNVs), segmental duplications or repetitive DNA sequences, and (iv) the GC content (guanine-cytosine) in each probe is from 19 to 80%.

As used in this application, the term long DNA probes refers to probes with a size in the range of 100 to 500 bp.

In one embodiment, the probes have a size in the range of 150 to 250 bp. In another embodiment, the probes have a size in the range of 160 to 180 bp.

In one embodiment of the method according to the present invention, long DNA probes span HSNRF.

In another embodiment of the method according to the present invention, the enriched cfDNA library comprises HSNRF.

In a preferred embodiment of the method according to the invention, the nucleic acid fragments are circulating cfDNA or RNA.

In one embodiment, the sample is a maternal plasma sample containing cell-free maternal DNA and cell-free fetal DNA (cffDNA).

The present invention can also be used with a variety of biological samples. Essentially any biological sample containing genetic material, such as RNA or DNA, and in particular cfDNA, can be used as a sample in the present invention. In one embodiment, the source of the DNA sample is a plasma sample containing cfDNA. In particular, for prenatal testing, the DNA sample contains fetal DNA (e.g., cffDNA). In one embodiment, for NIPT, the sample is a mixed sample that contains both maternal DNA and fetal DNA (e.g., cffDNA), such as a maternal plasma sample obtained from maternal peripheral blood. Typically, for mixed maternal/fetal DNA samples, the sample is a maternal plasma sample, although other tissue sources containing both maternal DNA and fetal DNA can be used. As used herein, the term mixed sample refers to a mixture of at least two biological samples originating from different sources, such as maternal/fetal DNA samples.

Depending on the conditions, the biological sample includes fetal DNA and maternal DNA, tumor-derived DNA and non-tumor-derived DNA, pathogen DNA or host DNA, and DNA originating from a transplanted organ and DNA originating from the host.

Thus, in the context of non-invasive diagnostics, the sample is a mixed sample, wherein said mixed sample is selected from the group comprising (i) fetal DNA and maternal DNA, (ii) tumor-derived DNA and tumor-derived DNA, (iii) pathogen DNA and host DNA, and (iv) DNA originating from a transplanted organ and DNA originating from the host.

Maternal plasma can be obtained from a peripheral whole blood sample of a pregnant subject, and the plasma can be obtained using standard methods. A total of 1-4 ml of plasma is sufficient to provide suitable DNA material for analysis according to the method described in the present invention. Total cfDNA can then be extracted from the sample using standard techniques, non-limiting examples of which include the QIAsymphony protocol (QIAGEN) suitable for the isolation of cffDNA, or any other manual or automated extraction method suitable for the isolation of cell-free DNA.

In the context of the present invention, the term subject refers to animals, preferably mammals and more preferably humans. The subject referred to in the present application is a pregnant subject and therefore preferably a female subject. A pregnant subject may be at any stage of pregnancy. The subject may become pregnant naturally or by artificial means. The term subject as used in the present application also refers to a subject suffering from a tumor or suspected of having a tumor. The above-mentioned subject may have undergone an organ transplant or be infected with a pathogen after transplantation or independently of the transplant.

To prepare biological samples, DNA is typically extracted using standard methods known in the art, a non-limiting example of which is the QIAsymphony protocol (QIAGEN).

Once cfDNA has been isolated, the sample is used to construct a sequencing library to make the sample compatible with subsequent sequencing technology such as next-generation sequencing (NGS). This typically involves ligating adapters to the ends of cfDNA fragments followed by amplification. Sequencing library preparation kits are commercially available or can be developed.

Preferably, in the method according to the present invention, the first and second nucleic acid fragments are selected from the groups comprising:

i) fetal DNA and maternal DNA, ii) tumor-derived DNA and non-tumor-derived DNA, iii) pathogen DNA and host DNA, iv) transplanted organ-derived DNA and host-derived DNA.

In the context of the present invention, the terms fetus and embryo are used interchangeably.

In one embodiment of the method, selecting a fraction of cfDNA fragments hybridizing to one or more probes spanning the HSNRF regions comprises the following steps:

(i) assignment of cfDNA fragments to the first and second cluster distributions, (ii) detection of hot spots of non-random fragmentation (HSNRF) using cfDNA fragments of the first cluster distribution, (iii) assignment of cfDNA fragments from the second cluster distribution to the third cluster, (iv) merging cfDNA fragments of the first cluster distribution with cfDNA fragments

- 3 047854 third cluster distribution.

As used in this application, the term cluster distribution refers to a subset (group) of cfDNA fragments having certain properties, such as, but not limited to, fragment length or fragment sequence. In the context of the present invention, cluster distribution is based on the size and sequence of fragments assigned to each group.

In one embodiment, the first cluster distribution comprises cfDNA fragments having a length less than or equal to 120 bp, or 125 bp, or 130 bp, or 135 bp, or 140 bp, or 145 bp, or 150 bp, or 155 bp, and the second cluster distribution comprises cfDNA fragments having a length greater than 120 bp, or 125 bp, or 130 bp, or 135 bp, or 140 bp, or 145 bp, or 150 bp, or 155 bp, respectively.

In another embodiment, the third cluster comprises selecting fragments of cfDNA from the second cluster whose ends overlap with the HSNRF detected in step (ii) above.

In one embodiment, preferably the probes span a hot spot of a non-random fragmentation site (HSNRF) such that only the 5' end of the fragmented nucleic acid is captured by the probe.

In another embodiment, the probes span the HSNRF site such that only the 3' end of extracellular nucleic acids derived from HSNRF can bind to the probe.

In another preferred embodiment, the probes span both HSNRF sites associated with the fragmented nucleic acid, such that the probe captures both the 5' and 3' ends of the extracellular nucleic acid associated with a given HSNRF site.

In another embodiment, combinations of the above are used.

Ideally, in the method of the present invention, the probes used to enrich cfDNA are double-stranded DNA fragments and (i) the length of each probe is 100-500 base pairs, (ii) each probe has a 5' end and a 3' end, (iii) preferably each probe binds to HSNRF at a distance of at least 10 base pairs, at both the 5' end and the 3' end, from regions bearing copy number variations (CNVs), segmental duplications or repetitive DNA sequences, and (iv) the GC content in each probe is from 19 to 80%.

In general, the hybridization step, preferably with probes as described above, can be performed before the sequencing library is generated or after the library is generated.

The region(s) of interest on the chromosome(s) of interest where the HSNRF is located is/are enriched by hybridizing a pool of probes to a sequencing library and then isolating those sequences from the sequencing library that bind to the probes. In one embodiment, the probe spans an HSNRF site such that only the 5' end of the fragmented extracellular nucleic acids is captured by the probe. In another embodiment, the probe spans an HSNRF site such that only the 3' end of the fragmented extracellular nucleic acids derived from the HSNRF can bind to the probe. In another preferred embodiment, the probe spans both HSNRF sites associated with the fragmented nucleic acid such that the probe captures both the 5' and 3' ends of the extracellular nucleic acid bound to the HSNRF site.

In order to facilitate the isolation of the desired enriched sequences (HSNRF), the probe sequences are typically modified such that the sequences that hybridize to the probes can be separated from the sequences that do not hybridize to the probes. This is typically accomplished by fixing the probes to a support. This allows for the physical separation of those sequences that bind the probes from those sequences that do not bind the probes. In one embodiment, each sequence in a pool of probes can be labeled with biotin, and the pool can then be bound to beads coated with a biotin-binding agent, such as streptavidin or avidin. In a preferred embodiment, the probes are labeled with biotin and bound to magnetic beads coated with streptavidin, allowing them to be separated by exploiting the magnetic properties of the beads. However, one of ordinary skill in the art will appreciate that other affinity binding systems exist in the art that can be used instead of the biotin-streptavidin/avidin system. For example, an antibody-based system can be used in which the probes are labeled with an antigen and then bind to the antibody-coated beads. Moreover, the probes can include a sequence marker at one end and can be bound to the support via a complementary sequence that hybridizes with such a sequence marker. Furthermore, in addition to magnetic beads, other types of supports can be used, such as polymer beads and the like.

In some embodiments, the sequencing library members that bind to the probe pool are fully complementary to the probes. In other embodiments, the sequencing library members that bind to the probe pool are partially complementary to the probes. For example, in certain circumstances, it is desirable to use and analyze data obtained from DNA fragments that are products of an enrichment process, but do not necessarily belong to

- 4 047854 regions of interest in the genome (for example, such DNA fragments may bind to probes due to partial homology (partial complementarity), and when sequenced will give very low coverage across the genome at coordinates other than the probes).

After enrichment of the sequence(s) of interest using probes, thereby generating a DNA library enriched in HSNRF sites, members of the enriched HSNRF library are eluted from the solid support, amplified, and sequenced using standard methods known in the art.

Next-generation sequencing (NGS) is typically used, although other sequencing technologies that provide very precise calculations in addition to sequence information can also be used. Accordingly, other methods of precise calculations such as digital polymerase chain reaction (PCR), single-molecule sequencing, nanopore sequencing, and microarrays can also be used instead of NGS.

In one embodiment of the method, the enriched sample is sequenced using a paired-end sequencing method, preferably a 2x75 bp sequencing method, to ensure that data on the start/end positions of the sequenced fragment is collected from both the 5' and 3' end, and to obtain sufficient sequencing information on the fragment to ensure de novo alignment/self-alignment.

In another embodiment, the sequenced fragments are grouped by size based on the degree of overlap or lack of overlap of their paired ends.

The invention relates to a method in which the nucleic acid fragment to be detected is present in the mixture at a concentration lower than the nucleic acid fragment from the same genetic locus but of a different origin. This means that if a specific locus is selected, for example, the maternal copy will be present 100 times and the fetal copy only once, in a solution containing the isolated cfDNA. In the case of cfDNA originating from the fetus, the component of the mixed sample originating from the fetus may have a range of possible values. For example, the range of fetal material in the mixed sample may be from 2 to 30%. Often, fetal fragments make up approximately 10% of the total DNA of the mixed sample. More importantly, in some compositions of mixed samples, the fetal DNA component of the sample may be less than 5%. In particular, in some compositions of samples, the material originating from the fetus makes up 3% or less of the total sample.

The present method is particularly suitable for the analysis of such low concentrations of target cfDNA. In the method according to the present invention, the nucleic acid fragment which is to be detected or the origin of which is to be determined and the nucleic acid fragment from the same genetic locus but of a different origin are present in a mixture in a ratio selected from the group: 1:2, 1:4, 1:10, 1:20, 1:50, 1:100, 1:200, 1:500, 1:1000, 1:2000 and 1:5000. The ratios are to be understood as approximate ratios, which implies deviations of plus/minus 30%, 20% or 10%. A person skilled in the art knows that such ratios will not have the exact numerical values specified above. These ratios are the ratio of the number of locus-specific molecules for a rare type to the number of locus molecules for a common type.

In one embodiment, the probes are provided in a form that allows them to bind to a support, such as biotinylated probes. In another embodiment, the probes are provided together with a support, such as biotinylated probes that are provided together with magnetic beads coated with streptavidin.

In a particular embodiment, the GC content of these or other probes is from 10 to 70%, preferably from 15 to 60%, more preferably from 20 to 50%.

In the described method, the result can be combined with additional statistical tests from a group including Student's t-test, two-dimensional nonparametric bootstrap test, stratified permutation test and fragment size proportion test.

In one embodiment, the method further comprises the step of combining statistical tests selected from the group including, but not limited to, a Student's t-test, a bivariate nonparametric bootstrap test, a stratified permutation test, ANOVA, any tests of proportions, and/or a regression model.

In the context of the present invention, the term partially refers to a region (section) of 10, 20, 30 or 40 bases of a nucleic acid fragment from the 5'-end or from the 3'-end. Therefore, the term used in this application refers entirely to a region (section) that covers 100% of the nucleic acid fragment. According to the method of the present invention, the probes hybridize to at least one site within the nucleic acid fragment. However, the probes can also target more than one site in the same nucleic acid fragment.

In one aspect of this application, probes for use in the method of the present invention are disclosed.

In another aspect of this application, a method is disclosed for use in diagnosing and/or screening a genetic abnormality in a sample.

- 5 047854

In one embodiment, the genetic abnormality is selected from the group including, but not limited to:

(a) aneuploidies of chromosomes 13, 18, 21 and/or X, Y;

(b) structural abnormalities, including, but not limited to, copy number changes, including microdeletions and microduplications, insertions, deletions, translocations, and small mutations, including point mutations.

In one embodiment, the method is used to detect epigenetic abnormalities, including but not limited to DNA and histone modifications. As used herein, the term epigenetic abnormalities refers to changes in gene expression with or without a change in DNA sequence. Accordingly, this term includes aneuploidies, microdeletions, microduplications, and point mutations, as well as changes in epigenetic modifications such as nucleotide methylation in DNA, or histone modifications such as histone acetylation or deacetylation, methylation, ubiquitination, phosphorylation, sumoylation, etc.

According to another aspect of the present invention, this application discloses a method for double enrichment of placenta-derived fragments in a cell-free DNA (cfDNA) mixture, comprising the following steps:

(a) obtaining a biological sample that contains a mixture of cell-free DNA (cfDNA) fragments, (b) preparing a sequencing library from the cfDNA fragments, (c) hybridizing the cfDNA library to a plurality of probes, wherein said probes preferably span at least one HSNRF, (d) isolating cfDNA fragments from the library that bind to the probes, (e) sequencing the cfDNA fragments from the library that bind to the probes, (f) removing duplicate sequenced reads, (g) selecting short cfDNA fragments, (h) detecting HSNRF in the short cfDNA fragments, (i) selecting long cfDNA fragments whose ends overlap with the HSNRF, (j) matching the selected cfDNA fragments from (g) and (i) to probe sequences.

In one embodiment, the short cfDNA fragments have a length less than or equal to 120 bp, or 125 bp, or 130 bp, or 135 bp, or 140 bp, or 145 bp, or 150 bp, or 155 bp, and the long cfDNA fragments have a length greater than 120 bp, or 125 bp, or 130 bp, or 135 bp, or 140 bp, or 145 bp, or 150 bp, or 155 bp, respectively.

In another embodiment, the cfDNA fragments from step (j) are used to detect aneuploidy in a fetus in a non-invasive prenatal diagnostic test.

In another embodiment, the cfDNA fragments from step (j) are used to detect a microdeletion/microduplication in a fetus in a non-invasive prenatal diagnostic test.

In another embodiment, the cfDNA fragments from step (j) are used to detect insertions, deletions, translocations, and small mutations, including point mutations, in the fetus.

In another aspect, the present invention provides a kit for performing the methods of the present invention, comprising:

a) probes that hybridize to at least one region in a nucleic acid fragment, wherein said at least one region partially or completely encompasses the nucleic acid fragment and, optionally,

b) reagents and/or software for implementing the determination and/or detection method.

In another aspect, the present application discloses a kit for performing a non-invasive test for use in the method of the present invention, wherein the kit comprises:

a) probes that hybridize to at least one region in nucleic acid fragments, said at least one region preferably encompassing at least one HSNRF, and, optionally,

b) reagents and/or software for performing the method described in aspects and embodiments of the present invention.

In one embodiment, the kit comprises a container consisting of a pool of probes, software, and instructions for performing the method.

In addition to the probe pool, the kit may contain (i) one or more components for isolating cfDNA from a biological sample, (ii) one or more components for preparing and enriching a sequencing library (e.g., primers, adapters, buffers, linkers, DNA modifying enzymes, ligation enzymes, probes, polymerase enzymes, etc.), (iii) one or more components for amplifying and/or sequencing the enriched library, and/or (iv) software for performing statistical analysis.

Accordingly, in another embodiment of the present invention, it is preferred that the fragment to be detected or determined has a smaller size than the second fragment. Depending on the tissue of origin, different fragment sizes are expected at the HSNRF locations. Since the cells of each tissue are differentiated to perform specific functions, different patterns of genetic activity are expected for each tissue type. Since gene activation depends on the tertiary structure of DNA, different tissues will have slightly different conformations, which leads to different cleavage sites and, therefore, different fragment sizes. For example, it has been demonstrated that cell-free DNA of placental origin is likely to be smaller in size compared to cell-free DNA of maternal origin.

Examples

The following examples illustrate the present invention and should not be construed as further limiting.

Example 1. Sample collection and library preparation.

A general methodology for double enrichment of placenta-derived DNA fragments in a mixed biological sample containing cell-free fetal and maternal DNA is explained for non-invasive prenatal diagnostics. In this example, methods for collecting and processing a maternal plasma sample (containing maternal and fetal DNA) are described. The same approach can be applied to other clinically applicable cases such as, but not limited to, oncology, genetic mutation, transplantation, and pathogen load assessment. In another aspect, the same approach can be applied to detect epigenetic modifications.

Collection of samples.

Plasma samples were obtained anonymously from pregnant women after the 10th week of pregnancy. The protocols used for sample collection were approved by the relevant bioethics committee and informed consent was obtained from all participants.

Sample extraction.

Cell-free DNA was extracted from the plasma of each individual using a manual or automated extraction method suitable for the isolation of cell-free DNA, such as, but not limited to, the QIAsymphony protocol suitable for the isolation of cell-free DNA (QIAGEN).

Preparation of sequencing library.

Isolated cell-free DNA from maternal plasma samples was used to construct a sequencing library. Initially, the 5' and 3' overhangs were repaired while the 5' ends were phosphorylated. The reaction products were purified using AMPure XP beads (Beckman Coulter). Sequencing adapters were then ligated to both ends of the DNA followed by purification using AMPure XP beads (Beckman Coulter). Single-strand breaks were removed in the polymerase reaction mixture and then purified using AMPure XP beads (Beckman Coulter). The library was amplified using a different polymerase enzyme (Koumbaris et al. (2016) Clinical Chemistry 62 (6): 848-855). The final library products were purified using AMPure XP beads (Beckman Coulter) and measured spectrophotometrically.

Example 2. Design and preparation of the probe.

This example describes the preparation of specific probes for the detection of HSNRF. The genomic target loci used to construct the probes were selected based on their GC content and their distance from repeat elements (minimum 50 bp). The probes can vary in size. In one embodiment of the method, the probes are between 100 and 500 bp in size and are generated using a PCR-based approach. The probes were prepared by simplex PCR using standard Taq polymerase, primers designed to amplify the target loci, and normal DNA as a template. In a preferred embodiment, the probe spans the HSNRF site such that only the 5' end of the fragmented nucleic acid is captured by the probe. In another embodiment, the probe spans the HSNRF site such that only the 3' end of extracellular nucleic acids derived from HSNRF can bind to the probe. In another preferred embodiment, the probe spans both HSNRF sites associated with the fragmented nucleic acid, such that the probe captures both the 5' and 3' ends of the extracellular nucleic acid associated with a given HSNRF site.

Example 3. Hybridization and probe amplification.

This example describes a method for targeted capture of nucleic acids by hybridization using probes, wherein the probes preferably span HSNRF, followed by quantification of the captured sequences using next generation sequencing (NGS).

Biotinylation of the probe.

Probes were prepared for blunt-end hybridization followed by purification. They were then ligated with a biotin adapter and purified. Probes were denatured before immobilization on streptavidin-coated magnetic beads.

- 7 047854

Hybridization of probes.

The amplified libraries were mixed with blocking oligonucleotides, Cot-1 DNA, salmon sperm DNA, hybridization buffer, blocking agent and denatured. Denaturation was followed by 30 min incubation at 37°C. The resulting mixture was then added to the biotinylated probes and incubated for 12-48 h at 60-70°C. After incubation, the enriched libraries were washed as described previously and the DNA was eluted by heating. The eluted products were amplified using externally linked adapter primers. The enriched amplified products were pooled equimolarly and sequenced on an appropriate platform.

Example 4. Bioinformatics analysis.

The inventors of the present invention have developed a computer-based method that organizes observed data (sequencing reads) into structures that allow subsets of sequenced reads to be selected from an enriched sample in order to increase the signal-to-noise ratio in a sample containing a mixture of cell-free DNA (cfDNA) fragments. The method according to the present invention groups and/or annotates a plurality of DNA sequences in such a way that the degree of homology between these DNA sequences is maximized if they actually originate from the same group and minimized otherwise. Specific sequence patterns are then identified in said data and used to assign them to specified regions of interest. In the present invention, the method is used for non-invasive detection of chromosomal abnormalities in a fetus, such as aneuploidies, microdeletions and microduplications in a sample containing a mixture of cell-free fetal DNA (cffDNA) and maternal DNA.

In a preferred embodiment, the FASTQ files of each sample are not aligned against a version of the human reference genome. Information about the fragment size and/or class is obtained using the length of individual reads (if greater than 150 bp) or information about the overlapping sequences of paired-end reads.

Identical reads are removed and either de novo assembly or alignment to pre-defined short sequence targets is performed.

The average person skilled in the art will appreciate the many readily available and well-established tools for performing sequence matching.

The algorithm is a set of data processing procedures, mathematical and statistical models, organized as a series of steps. A step of the algorithm is aimed at determining the relative copy number of the chromosome of interest and/or the subchromosomal region of interest in the enriched sample and is used to detect complete or partial chromosomal abnormalities, such as aneuploidies of chromosomes 13, 18, 21, X, Y or any other chromosome, as well as microdeletion/microduplication syndromes and other small mutations.

The key feature of the method is the double enrichment of the sample for the fetal (placental) fraction by selecting a subset of enriched fragments, which leads to an increase in the signal-to-noise ratio (fetal fraction to maternal fraction, e.g., the proportion of fragments of placental origin). To this end, the developed method identifies and applies fragments that are highly likely to originate from the placenta. Fig. 3 shows the predicted size distribution of fragments selected by the claimed method compared to the size distribution of all fragments of the representative sample, demonstrating the enrichment of shorter fragments that are mainly derived from fetal (placental) tissues.

In one embodiment of the method, the enriched sample is sequenced using a 2x75 bp sequencing method. In this embodiment, the sequenced fragments are grouped by size based on the degree of overlap or lack of overlap of their paired-end reads. Each fragment is classified into two groups, respectively: short (Group 1) and long (Group 2). HSNRFs are then identified using the sequence similarity of at least the 20 most distal nucleotides of the fragments in Group 1. Fragments from Group 2 whose ends overlap with the identified HSNRFs are then used to create Group 3. All of the fragments from Group 1 and all of the fragments from Group 3 are then retained for subsequent analysis. These Group 1 cfDNA fragments are smaller in size and enriched in cell-free fetal DNA. The reported Group 3 cfDNA fragments are larger in size but are also enriched in cell-free fetal DNA because their ends overlap with HSNRF. Group 2 cfDNA fragments whose ends do not overlap with HSNRF have a higher probability of maternal origin and are excluded from all subsequent analyses. Figure 3 shows the size distribution of the fragments selected by the claimed method (Group 1 and Group 3) compared to the size distribution of all fragments in a typical sample, demonstrating enrichment in fetal (placental) cfDNA fragments.

After matching all fragments from Group 1 and Group 3 to the probe sequences for each test sample, a classification score is calculated by comparing the number of remaining fragments on the target chromosome, e.g. chromosome 21 (Fig. 1) and/or the target subchromo-

Claims (6)

somatic region, for example, 22q11.2 (Fig. 2) with the number of fragments on reference chromosomes and/or subchromosomal regions. In another embodiment of the method, the said steps can be carried out after aligning the sequenced reads with the reference genome. CLAUSE OF INVENTION 1. A method for detecting a chromosomal abnormality in a mixed sample, including the following steps: (a) obtaining a biological sample, said sample comprising a mixture of cell-free DNA fragments (cfDNA), (b) preparing a sequencing library from said cfDNA fragments, (c) hybridizing one or more probes to at least one or more cfDNA fragments, wherein said one or more probes span at least one genomic region comprising, at a distance of less than 300 bp, genomic bases distinguishing two tissue types present in the cfDNA mixture (HSNRF), (d) isolating cfDNA fragments from said library that bind to said probes, (e) sequencing said cfDNA fragments from the library that bind to the probes, (f) using information about the size, start and/or end of each or a subset of enriched cfDNA fragments from steps (c-e) to select a fraction of cfDNA fragments hybridized to said one or more probes, wherein step (f) further comprises the following steps: (i) assigning the cfDNA fragments to a first and a second subset, wherein said first subset comprises cfDNA fragments having a length less than or equal to 145 bp, and said second subset comprises cfDNA fragments having a length greater than 145 bp, (ii) detecting at least one genomic region containing, at a distance of less than 300 bp, genomic bases that distinguish between two tissue types present in a mixture of cfDNA (HSNRF) using cfDNA fragments of a first subset, (ii i) assigning cfDNA fragments from the second subset to a third subset, wherein said third subset comprises selecting cfDNA fragments from said second subset whose ends overlap with the HSNRF detected in step (ii), (iv ) combining the cfDNA fragments of the first subset with the cfDNA fragments of the third subset, (v) matching all of the cfDNA fragments from the first and third subsets to probe sequences, and (vi ) calculating a classification score based on the cfDNA fragments according to (v) by comparing the number of cfDNA fragments remaining on the target chromosome on which the abnormality is tested with the number of cfDNA fragments on the reference chromosome, wherein the chromosomal abnormality is detected when the classification score above the threshold value of 1. 2. The method according to claim 1, characterized in that said probes are long DNA molecules and: (i) the length of each probe is from 100 to 500 base pairs, (ii) each denatured probe has a 5' end and a 3' end, (iii) preferably, each probe binds to HSNRF at a distance of at least 10 base pairs at both the 5' end and the 3' end from regions bearing copy number variations (CNVs), segmental duplications or repetitive DNA elements, and (iv) the GC content of each probe is from 19 to 80%. 3. The method according to any one of claims 1, 2, characterized in that said sample is a mixed sample selected from the group comprising: (i) embryonic DNA and maternal DNA, (ii) tumor-derived DNA and non-tumor-derived DNA, (iii) pathogen DNA and host DNA, and (iv) DNA originating from a transplanted organ and DNA originating from the host. 4. The method according to any one of paragraphs 1-3, characterized in that said method additionally includes the step of combining statistical tests selected from a group containing, but not limited to, a t-test, a two-dimensional nonparametric bootstrap test, a stratified permutation test, a nonparametric test, ANOVA, and a fragment size proportion test. 5. The method according to any one of claims 1 to 4 for use in diagnosing and/or screening a genetic abnormality in a sample. 6. The method according to paragraph 5, characterized in that the said genetic anomaly is selected from a group containing, but not limited to, the following: (a) aneuploidies of chromosomes 13, 18, 21 and/or X, Y, (b) structural abnormalities, including, but not limited to, copy number changes, including microdeletions and microduplications, insertions, deletions, translocations and small mutations -