RU2825664C2 - Sequence graph tool for determining variations in regions of short tandem repeats - Google Patents

Abstract

FIELD: biotechnology; medicine.

SUBSTANCE: described is a method for sequencing genomic loci comprising two or more repeat sequences. Method is realized using a computer equipped with one or more processors and system memory for genotyping one or more repeat sequences. Each of the sequences contains one or more sub-sequences of repeats. Method includes: obtaining a sequence graph, where the sequence graph has a graph data structure with vertices representing nucleotide sequences, and the directed edges connect the vertices, and wherein the sequence graph contains two or more proper simple cycles, wherein each proper simple cycle is a subsequence of repeats; alignment, using one or more processors, of reading sequences of the analyzed sample on a reference genome to determine genomic coordinates of reading sequences and selecting a subset of reads of sequences; and aligning, using one or more processors, a selected subset of reads of sequences with two or more sequences of repetitions, represented by a sequence graph representing a genomic locus. Also disclosed is a corresponding system for sequencing genomic loci, including two or more sequences of repeats.

EFFECT: invention enables to genotype sequences of repeats, including short tandem repeats (CTR), which are significant from a medical point of view.

10 cl, 7 dwg, 1 tbl

Description

INCORPORATION BY LINK

[0001] A PCT Request Form is filed concurrently with this specification in this application. Each application for which benefit or priority is claimed in this application as indicated in a concurrently filed PCT Request Form is incorporated herein by reference in its entirety and for all purposes.

PREREQUISITES FOR THE CREATION OF THE INVENTION

[0002] Repeat expansions are a special class of microsatellite and minisatellite variants involving short tandem repeat (STR) polymorphisms. Repeat expansions are also called dynamic mutations due to their instability, where the STRs expand beyond a certain size. Genetic diseases caused by unstable repeat expansions include, but are not limited to, fragile X syndrome (FXS), Huntington's disease, and amyotrophic lateral sclerosis (ALS).

[0003] Detection of repeat expansions is important for the diagnosis and treatment of certain genetic diseases. However, it is difficult to determine repeat sequences using short reads that do not completely cover the repeat sequence. Thus, it is desirable to develop methods that use short reads to detect medically significant repeat expansions.

SUMMARY OF THE INVENTION

[0004] The disclosed embodiments relate to methods, devices, systems, and computer program products for sequencing genomic loci, including repeat sequences, including short tandem repeat sequences, which may be associated with genetic disorders. The methods include aligning reads to sequence graphs, each of which represents a genomic locus, and using the aligned reads to genotype one or more repeat sequences at the genomic locus. The sequence graphs are directed graphs, each of which includes at least one proper simple cycle representing a repeat sequence.

[0005] In a first aspect of the description, methods are provided for genotyping genomic loci using a computer, including repeat sequences. The methods are implemented using a computer including one or more processors and a system memory. These methods can be used to genotype one or more repeat sequences, each of which comprises one or more repeat subsequences. The methods include: (a) collecting sequence reads of a test sample from a database using one or more processors; (b) aligning, using the one or more processors, the reads of the one or more repeat sequences, each of which is represented by a sequence graph, wherein the sequence graph has a directed graph data structure, wherein the nodes represent nucleotide sequences and directed edges connect the nodes, and wherein the sequence graph contains one or more proper simple cycles, wherein each proper simple cycle represents a repeat subsequence, wherein each repeat subsequence contains repeats of a repeating unit of one or more nucleotides; and (c) determining, by one or more processors, one or more genotypes of one or more repeat sequences using sequence reads aligned to the one or more repeat sequences.

[0006] In some embodiments, a repeat sequence of one or more repeat sequences comprises a specific repeat unit comprising at least one partially defined nucleotide. In some embodiments, the specific repeat unit comprises degenerate codons.

[0007] In some embodiments, one or more of the native simple cycles comprise two or more native simple cycles representing two or more repeat subsequences.

[0008] In some embodiments, the sequence graph further comprises two or more alternative pathways for the two or more alleles. In some embodiments, the two or more alleles comprise a deletion or a substitution. In some embodiments, the substitution comprises a single nucleotide variant (SNP) or a single nucleotide polymorphism (SNP). In some embodiments, the method further comprises genotyping the two or more alleles using sequence reads aligned to the two or more alternative pathways. In some embodiments, genotyping the two or more alleles comprises applying the coverage of the two or more alternative pathways to a probability model to determine the probabilities of the two or more alleles. In some embodiments, the probability model models the probability of an allele as a function of the coverage of the allele, wherein the function is selected from a Poisson distribution, a negative binomial distribution, a binomial distribution, or a beta-binomial distribution. In some embodiments, the rate parameter of the Poisson distribution is estimated from the read length and the average depth observed at the genomic locus.

[0009] In some embodiments, the method further comprises aligning, prior to (b), the sequence read to a reference genome to determine genomic coordinates of the sequence read, and selecting a subset of the sequence read as the sequence read aligns to one or more repeat sequences, each of which is represented by a sequence graph. In some embodiments, the subset of sequence reads includes reads aligned to the region represented by the sequence graph or a space near it. In some embodiments, the subset of sequence reads includes unaligned reads whose mate reads map to the region represented by the sequence graph or a space near it. In some embodiments, the subset of sequence reads comprises a sequence read aligned to one or more off-target regions that are known hotspots for misaligned reads.

[0010] In some embodiments, aligning a sequence read with a sequence graph includes: finding a kmer correspondence between the sequence read and a sequence graph path; and extending the kmer correspondence to a complete alignment of the nodes and edges of the sequence graph, including one or more proper simple cycles.

[0011] In some embodiments, aligning a sequence read to a sequence graph includes pruning the graph by removing the ends of low confidence alignments.

[0012] In some embodiments, aligning a sequence read to a sequence graph includes combining alignments by: aligning subsequences of the read to the sequence graph; and combining the subsequence alignments to fully align the sequence read.

[0013] In some embodiments, the method further includes generating a sequence graph based on a locus specification including the structure of the genomic locus.

[0014] In some embodiments, the sequence reads include paired end reads, and operation (c) includes the following components: (i) determining base and anchored reads in the paired end reads, wherein the base reads are aligned to one or more repeat sequences or to a space adjacent thereto, and the anchored reads are unaligned reads that are paired with the anchored reads; and (ii) determining one or more genotypes for the one or more repeat sequences using at least the anchored reads.

[0015] In some embodiments, operation (ii) includes determining one or more genotypes for one or more repeat sequences using base reads as well as anchored reads. In some embodiments, base reads are aligned to within about 5 kb of the repeat sequence. In some embodiments, unaligned reads include reads that cannot be aligned or are poorly aligned with the sequence graph.

[0016] In some embodiments, one or more repeat sequences comprise a short tandem repeat (STR) sequence. In some embodiments, the STR expansion is associated with fragile X syndrome, amyotrophic lateral sclerosis (ALS), Huntington's disease, Friedreich's ataxia, spinocerebellar ataxia, spinobulbar muscular atrophy, myotonic dystrophy, Machado-Joseph disease, or dentato-rubro-pallido-Lewis atrophy.

[0017] In some embodiments, the method further includes using a sequencer to generate paired end reads of the sample under study.

[0018] In some embodiments, the method further includes extracting the test sample from the subject.

[0019] In some embodiments, the test sample is a blood, urine, saliva, or tissue sample.

[0020] In some embodiments, the repeating unit comprises from 1 to 50 nucleotides.

[0021] In some embodiments, the read is shorter than at least one of the one or more repeat sequences.

[0022] In another aspect of the disclosure, systems are provided for genotyping genomic loci, including repeat sequences. In some embodiments, a system includes: a system memory and one or more processors configured to perform the following actions: (a) collecting, using the one or more processors, sequence reads of a test sample from a database; (b) aligning, using the one or more processors, the reads of one or more repeat sequences, each of which is represented by a sequence graph, wherein the sequence graph has a directed graph data structure, wherein the nodes represent nucleotide sequences and directed edges connect the nodes, and wherein the sequence graph comprises one or more proper simple cycles, wherein each proper simple cycle represents a subsequence of the repeats, wherein each subsequence of the repeats comprises repeats of a repeat unit of one or more nucleotides; and (c) determining, using the one or more processors, one or more genotypes of the one or more repeat sequences using the sequence reads aligned to the one or more repeat sequences.

[0023] In some embodiments, the system also includes a sequencer for sequencing nucleic acids of the test sample.

[0024] In some embodiments, one or more processors are configured to operate in accordance with the various methods described herein.

[0025] In another aspect of the description, a computer program product is provided, including a computer-readable medium intended for long-term storage of information, containing program code, execution of which by one or more processors of the computer system causes the computer system to implement the above-described methods for genotyping genomic loci, including repeat sequences. The program code includes (a) code for collecting sequence reads of a test sample from a database; (b) code for aligning the sequence reads with one or more repeat sequences, each of which is represented in a sequence graph, wherein the sequence graph has a data structure of a directed graph, wherein the nodes represent nucleotide sequences and directed edges connecting the nodes, and wherein the sequence graph contains one or more proper simple cycles, wherein each proper simple cycle represents a subsequence of repeats, wherein each subsequence of repeats contains repeats of a repeating unit of one or more nucleotides; and (c) code for determining one or more genotypes of one or more repeat sequences using sequence reads aligned to the one or more repeat sequences.

[0026] In some embodiments, the program code includes code for performing operations within the methods described herein.

[0027] Although the examples and terminology used herein relate to the human body, the concepts described herein are applicable to the genomes of any plant or animal. An understanding of these and other objects and features of the present disclosure is aided by the description below and the appended claims; in addition, the necessary information may be obtained by practicing the contents of the description below.

BRIEF DESCRIPTION OF GRAPHIC MATERIALS

[0028] Fig. 1A is a schematic diagram illustrating the difficulties of aligning a sequence with a repeat sequence on a reference sequence.

[0029] Fig. 1B is a schematic diagram illustrating alignment of sequence reads using paired end reads in accordance with certain embodiments to overcome the difficulties illustrated in Fig. 1A.

[0030] Fig. 1C is a flow chart illustrating a process for genotyping a genomic locus that includes a repeat sequence, in accordance with some embodiments.

[0031] Fig. 1D illustrates a first sequence graph representing a first genomic locus.

[0032] Fig. 1E illustrates a second sequence graph representing a second genomic locus.

[0033] Fig. 1F illustrates a third sequence graph representing a third genomic locus.

[0034] Fig. 1G is a schematic diagram of a process for determining genotypes of HTT locus variants that include two CTP sequences, in accordance with some embodiments.

[0035] Fig. 1H is a schematic diagram of the process for determining genotypes of variants at the Lynch I locus, including the ONV and the CTP, according to some embodiments. The left panel of Fig. 1H is a schematic diagram of the general process for performing targeted genotyping; the right panel illustrates the application of this process to genotyping variants at the locus associated with Lynch I syndrome.

[0036] Fig. 2 is a block diagram illustrating a high level view of an example of a method for determining the presence or absence of a repeat sequence expansion in a sample.

[0037] Figs. 3 and 4 are flow charts illustrating examples of methods for detecting repeat expansion using paired end reads.

[0038] Fig. 5 is a flow chart of a method in which unaligned reads that are not associated with any repeat sequence of interest are used to determine repeat expansion.

[0039] Fig. 6 shows a block diagram of a dispersed system for processing a test sample.

[0040] Fig. 7 illustrates the genotyping accuracy of CAG and CCG TTPs at the HTT locus based on simulated data using different methods.

DETAILED DESCRIPTION

[0041] The disclosure relates to methods, devices, systems, and computer program products for determining target repeat expansions, including repeat sequence expansions, that are medically significant. Examples of repeat expansions include, but are not limited to, expansions associated with genetic disorders such as fragile X syndrome, ALS, Huntington's disease, Friedreich's ataxia, spinocerebellar ataxia, spinobulbar muscular atrophy, myotonic dystrophy, Machado-Joseph disease, and dentato-rubro-pallido-Lewis atrophy.

[0042] Unless otherwise indicated, the practice of the methods and systems described herein involves standard techniques and apparatus commonly used in molecular biology, microbiology, protein purification, protein engineering, protein and DNA sequencing, and recombinant DNA fields, which are within the skill of the art. Such techniques and apparatus are known to those skilled in the art and are described in numerous reference publications (see, e.g., Sambrook et al., “Molecular Cloning: A Laboratory Manual,” Third Edition (Cold Spring Harbor), [2001]); and Ausubel et al., “Current Protocols in Molecular Biology” [1987]).

[0043] Numerical ranges include the numbers defining the range. Each maximum numerical limitation appearing in this specification is intended to include every lesser numerical limitation, as if such lesser numerical limitations were expressly written herein. Each minimum numerical limitation appearing in this specification will include every greater numerical limitation, as if such greater numerical limitations were expressly written herein. Each numerical range appearing in this specification will include every narrower numerical range that is within such broader numerical range, as if all such narrower numerical ranges were expressly written herein.

[0044] The headings provided in this document are not intended to be limiting with respect to the present description.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. A variety of scientific terminology, including terminology used herein, is well known and available to those skilled in the art. Although any methods and materials similar or equivalent to those described herein find use in the practice or testing of the embodiments described herein, certain methods and materials have been described.

[0046] The terms defined below are more fully described by reference to the full version of the specification. It should be understood that this description is not limited solely to specific methodology, protocols, and reagents, as these may vary depending on the context in which they are used by those skilled in the art.

Definitions

[0047] As used herein, the singular terms include the plural forms of the same terms unless the context otherwise requires.

[0048] Unless otherwise noted, nucleic acids are written left to right in a 5' to 3' orientation, and amino acid sequences are written left to right in an amino to carboxy orientation, respectively.

[0049] The term "plurality" refers to more than one element. For example, the term is used herein to refer to a number of nucleic acid molecules or sequence reads sufficient to detect significant differences in repeat expansions between test samples and control samples using the methods described herein.

[0050] The term "repeat sequence" refers to a nucleotide sequence that includes repeated inclusions of a shorter sequence. As used herein, the shorter sequence is referred to as a "repeat unit." Repeated instances of inclusion of repeat units are referred to as "repeats" or "copies" of the repeat unit. In many contexts, the location of the repeat sequence is associated with a gene encoding a protein. In other situations, the repeat sequence may be in a non-coding region. Repeat units may be formed in a repeat sequence with or without gaps between repeat units. For example, in normal samples, the FMR1 gene typically includes an AGG gap in CGG repeats, such as (CGG)10 + (AGG) + (CGG)9. Gapless patterns, as well as long repeat sequences with few gaps, are prone to expansion of the associated gene repeats, which can lead to genetic diseases because the repeats expand beyond a certain number. In various embodiments of the present disclosure, the number of repeats is counted in the same way as the number of in-frame repeats, regardless of the presence of gaps. Methods for assessing in-frame repeats are further described below.

[0051] In various embodiments, the repeat units comprise from 1 to 100 nucleotides. Many well-characterized repeat units are trinucleotide or hexanucleotide units. Some other repeat units that have been well characterized and are applicable to the embodiments described herein include, but are not limited to, units of 4, 5, 6, 8, 12, 33, or 42 nucleotides. See, for example, Richards (2001) Human Molecular Genetics, Vol. 10, No. 20 , 2187-2194. The scope of the present invention is not limited to the specific number of nucleotide bases described above, so long as they are relatively short compared to a repeat sequence having multiple repeats or copies of the repeat units. For example, the repeating unit may comprise at least 3, 6, 8, 10, 15, 20, 30, 40, 50 nucleotides. Alternatively or additionally, the repeating unit may comprise no more than about 100, 90, 80, 70, 60, 50, 40, 30, 20, 10, 6 or 3 nucleotides.

[0052] A repeat sequence can expand under developmental and mutagenesis conditions to produce more copies of the same repeat unit. This is referred to in the art as "repeat expansion." This process is also referred to as "dynamic mutation" due to the unstable nature of the repeat unit expansion. Some repeat expansions have been shown to be associated with genetic diseases and pathological symptoms. Other repeat expansions are poorly understood. The methods described herein can be used to identify both known and novel repeat expansions. In some embodiments, a repeat sequence having a repeat expansion is longer than about 100, 150, 300, or 500 nucleotide base pairs (bp). In some embodiments, a repeat sequence having a repeat expansion is greater than about 1,000 bp. o., 2,000 bp, 3,000 bp, 4,000 bp, 5,000 bp or 10,000 bp, etc.

[0053] In graph theory, a vertex and an edge are the two basic units from which graphs are constructed. A vertex or node is one of the points in a graph that can be connected by edges. In a graph diagram, a vertex can be represented as a shape with a label, and an edge can be represented as a line (an undirected edge) or an arrow (a directed edge) running from one vertex to another.

[0054] Two vertices connected by an edge are considered to be endpoints of the edge. A vertex × is considered adjacent to another vertex y if the graph contains the edge (x, y).

[0055] An undirected graph consists of a set of vertices and a set of undirected edges (connecting unordered pairs of vertices), while a directed graph consists of a set of vertices and a set of directed edges (connecting ordered pairs of vertices).

[0056] In graph theory, each edge has two vertices (in hypergraphs there may be more) to which it is attached, called its endpoints. Edges may be directed or undirected; undirected edges are also called lines, and directed edges are also called arcs or arrows.

[0057] A directed edge is an edge that connects a top and bottom vertex, where the top vertex is before the directed edge and the bottom vertex is after the directed edge.

[0058] An undirected edge is an edge that connects two vertices, where either vertex may be in front of the other on a path in the graph.

[0059] In this document, the terms "cycle", "proper simple cycle", and "single-node cycle" are used interchangeably. A cycle has one node and an edge, both ends of which are connected to one node.

[0060] A cycle is a path containing two or more vertices, where the path of the cycle begins and ends with the same vertex. A simple cycle is a cycle that has no repeating vertices or edges other than from the start and end vertices.

[0061] A cyclic graph is a graph that contains at least one cycle.

[0062] An acyclic graph is a graph that does not contain cycles or proper simple cycles.

[0063] A directed acyclic graph (DAG) is a directed graph without any cycles or proper simple cycles.

[0064] A path in a graph is a sequence of vertices and edges in which both endpoints of an edge are adjacent to an edge in the sequence. A path in a directed graph has a top vertex that is before the directed edge (or arc/arrow) and a bottom vertex that is after the directed edge.

[0065] The Poisson distribution is a discrete probability distribution that expresses the probability of a given number of events occurring within a fixed interval of time or space if those events occur at a known constant rate and regardless of the time since the last event.

[0066] Fully defined base symbols include G, A, T, C to represent guanine, adenine, thymine, and cytosine, respectively.

[0067] The list of partially defined nucleic acids includes, among other things, the following components:

[0068] Purine (adenine or guanine): R

[0069] Pyrimidine (thymine or cytosine): Y

[0070] Adenine or thymine. W

[0071] Guanine or cytosine: S

[0072] Adenine or cytosine: M

[0073] Guanine or Thymine: K

[0074] Adenine, thymine, or cytosine: H

[0075] Guanine, cytosine or thymine: B

[0076] Guanine, adenine or cytosine: V

[0077] Guanine, adenine, or thymine. D

[0078] Guanine, adenine, thymine, or cytosine: N

[0079] The term "paired-end reads" refers to reads obtained by paired-end sequencing, which produces one read from each end of a nucleic acid fragment. Paired-end sequencing involves fragmenting DNA to form sequences called inserts. In some protocols, including those used by Illumina, reads from shorter inserts (e.g., tens to hundreds of bp) are referred to as paired-end reads with short inserts or paired-end reads. In contrast, reads from longer inserts (e.g., on the order of several thousand bp) are referred to as mate-paired reads. As used herein, paired-end reads with short inserts and mate-paired reads with long inserts may be used; however, they are not differentiated with respect to the repeat expansion analysis process. Therefore, the term "paired end reads" can refer to both short insert paired end reads and long insert mate pair reads, which are further described below herein. In some embodiments, paired end reads include reads from about 20 bp to 1,000 bp. In some embodiments, paired end reads include reads from about 50 bp to 500 bp, from about 80 bp to 150 bp, or about 100 bp. It should be understood that the two paired end reads need not be located at the extreme end of the fragment being sequenced. Instead, one or both reads may be located near the end of the fragment. Moreover, the methods exemplified herein in the context of paired end reads can be performed with any of a plurality of paired end reads, regardless of whether the reads are obtained from the end or another part of the fragment.

[0080] As used herein, the term "alignment" refers to the process of comparing a read to a reference sequence and determining whether the reference sequence contains the read sequence. The alignment process attempts to determine whether a read can be aligned with the reference sequence, but does not always result in a read that is aligned with the reference sequence. If the reference sequence contains a read, the read may align with the reference sequence or, in some embodiments, with a specific location in the reference sequence. In some cases, an alignment simply indicates whether a read is a member of a specific reference sequence (i.e., whether the read is present in the reference sequence or not). For example, aligning a read to a reference sequence for human chromosome 13 will indicate whether the read is present in the reference sequence for chromosome 13. A tool that provides this information may be referred to as a set membership testing module. In some cases, the alignment further specifies the location in the reference sequence to which the read is aligned. For example, if the reference sequence is the human whole genome sequence, the alignment may indicate that the read is on chromosome 13 and may further specify that the read is on a specific strand and/or site on chromosome 13.

[0081] Aligned reads are one or more sequences that are determined to match, according to the order of their nucleic acid molecules, a known reference sequence, such as a reference genome. An aligned read and its determined location on the reference sequence constitute a sequence tag. Alignment can be performed manually, although it is typically implemented using a computer algorithm, since it is not possible to align reads in a reasonable amount of time to implement the methods described herein. One example of a sequence alignment algorithm is the Efficient Local Alignment of Nucleotide Data (ELAND) computer program distributed as part of the Illumina genomic analysis workflow. In an alternative embodiment, a Bloom filter or similar set membership testing module can be used to align reads to reference genomes. See U.S. Patent Application No. 14/354,528, filed April 25, 2014, which is incorporated herein by reference in its entirety. The alignment of the sequence reads in the alignment process may be a 100% or less than 100% sequence match (i.e., a non-perfect match).

[0082] As used herein, the term "alignment" means assigning a sequence of reads to a larger sequence, such as a reference genome, by alignment.

[0083] In some cases, a read at one end of two paired-end reads is aligned with a repeat sequence of a reference sequence, and a read at the other end of the two paired-end reads is not aligned. In such cases, the paired-end read that is aligned with the repeat sequence of the reference sequence is called a "base read." A paired-end read that is not aligned with the repeat sequence but is linked to the base read is called an anchored read. Thus, a misaligned read can be anchored and linked to the repeat sequence. In some embodiments, misaligned reads include both reads that cannot be aligned with the reference sequence and reads that are poorly aligned with the reference sequence. If a read is aligned with the reference sequence and the mismatch level is above a certain criterion, the read is considered poorly aligned. For example, in various embodiments, a read is considered poorly aligned if at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 mismatches were noted during its alignment. In some cases, both reads of a pair are aligned with a reference sequence. In such cases, both reads may be analyzed as "base reads" in various embodiments.

[0084] The terms "polynucleotide,""nucleicacid," and "nucleic acid molecules" are used interchangeably and refer to a covalently linked sequence of nucleotides (e.g., ribonucleotides for RNA and deoxyribonucleotides for DNA) in which the 3' pentose position of one nucleotide is linked by a phosphodiester group to the 5' pentose position of the next nucleotide. Nucleotides include sequences of any form of nucleic acid, including, but not limited to, RNA and DNA molecules, such as cell-free DNA (cfDNA). The term "polynucleotide" includes, but is not limited to, single- and double-stranded polynucleotides.

[0085] The term "test sample" as used herein refers to a sample, typically obtained from a biological fluid, cell, tissue, organ, or organism, that contains a nucleic acid or mixture of nucleic acids that contains at least one nucleotide sequence that is to be tested for copy number variation. In certain embodiments, the sample has at least one nucleotide sequence that is suspected of having a copy number variation. Such samples include, but are not limited to, sputum/oral fluid, amniotic fluid, blood, blood fraction, or fine needle aspiration biopsy samples, as well as urine, peritoneal fluid, pleural fluid, etc. Although the sample is often collected from a human subject (e.g., a patient), the assays can be used to analyze copy number variations (CNVs) in samples collected from any mammal, including, but not limited to, dogs, cats, horses, goats, sheep, cattle, pigs, etc. The sample can be used directly as obtained from the biological source or after pre-treatment to alter the nature of the sample. For example, such pre-treatment may include obtaining plasma from blood, diluting viscous liquids, etc. Pre-treatment methods may also include, without limitation, filtration, precipitation, dilution, distillation, mixing, centrifugation, freezing, lyophilization, concentration, amplification, fragmentation of nucleic acids, inactivation of interfering components, addition of reagents, lysing, etc. When such pre-treatment methods are used with respect to a sample, such pre-treatment methods typically imply that the nucleic acid(s) remains(remain) in the test sample, sometimes at a concentration that is proportional to their concentration in the untreated test sample (e.g., in a sample that is not subjected to such type(s) of pre-treatment). Such "treated" samples are still considered biological "test" samples with respect to the methods described herein.

[0086] A control sample may be either negative or positive. The term "negative control sample" or "unaffected sample" refers to a sample containing nucleic acids that are known or expected to have a repeat sequence with a repeat number in a range that is not pathogenic. A "positive control sample" or "affected sample" is known to have a repeat sequence with a repeat number in a range that is pathogenic. Repeats in the repeat sequence of a negative control sample typically have not expanded beyond the normal range, whereas repeats in the repeat sequence of a positive control sample typically have expanded beyond the normal range. In this way, nucleic acids in a test sample can be compared to one or more control samples.

[0087] The term "sequence of interest" as used herein refers to a nucleotide sequence associated with differences in sequence presentation between healthy and diseased patients. The sequence of interest may be a repeat sequence on a chromosome that is expanded in a disease, including a hereditary disease. The sequence of interest may be a portion of a chromosome, a gene, a coding sequence, or a non-coding sequence.

[0088] As used herein, the term "next-generation sequencing (NGS)" refers to sequencing methods that enable massively parallel sequencing of clonally amplified molecules and individual nucleic acid molecules. Non-limiting examples of NGS include sequencing by synthesis using reverse dye terminators and sequencing by ligation.

[0089] The term "parameter" as used herein refers to a numerical value that characterizes a physical property. Often, a parameter is used to numerically characterize a set of quantitative data and/or a numerical relationship between sets of quantitative data. For example, a parameter is a ratio (or ratio function) between the number of sequence tags mapped to a chromosome and the length of the chromosome to which the tags are mapped.

[0090] The term "discrimination criterion" as used herein refers to any number or quantity that is used as a cutoff level for characterizing a sample, such as a test sample, that contains a nucleic acid and is collected from a subject suspected of having a medical condition. The cutoff value can be compared to a parameter value to determine whether a sample that yields such a parameter value indicates the presence of a medical condition in the subject. In certain embodiments, the cutoff value is calculated using a set of control data and is used as a cutoff when diagnosing a repeat expansion in an organism. In some embodiments, if the cutoff value is exceeded by the results obtained by the methods described herein, the subject can be diagnosed with a repeat expansion. In the methods described herein, appropriate cutoff values can be determined by analyzing values calculated relative to a training set of samples or control samples. Threshold values can also be calculated based on empirical parameters such as sequencing depth, read length, repeat sequence length, etc. Alternatively, affected samples that are known to have repeat expansion can also be used to confirm that the selected threshold values can be used to distinguish between unaffected samples within the test sequence. The choice of threshold value depends on the level of confidence that the user wishes to make for the classification. In some embodiments, the training set used to determine the corresponding threshold values contains at least 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 2,000, 3,000, 4,000 or more qualified samples. The use of larger sets of qualified samples to increase the diagnostic value of cut-off values may be preferable.

[0091] The term "read" refers to a reading of the sequence of a portion of a nucleic acid sample. Typically, a read is a short sequence of coherent nucleotide base pairs in a sample. A read may symbolically represent the sequence of nucleotide base pairs (in the form of ATCG) of a portion of a sample. It may be stored in a memory device and processed as needed to determine whether it matches a reference sequence or other criteria. A read may be obtained directly by a sequencing machine or indirectly from stored sequence information related to a sample. In some cases, a read is a DNA sequence of sufficient length (e.g., at least about 25 bp) that can be used to determine a larger sequence or region, such as one that can be aligned and compared to a chromosome, genomic region, or gene.

[0092] The term "genome sequencing" is used to refer to sequencing of any segment of a subject's entire genome.

[0093] The term "site" refers to a unique location (i.e., chromosome identifier, chromosomal location, and orientation) in a reference genome. In some embodiments, a site may be a residue, a sequence tag, or a segment location on a sequence.

[0094] As used herein, the term "reference genome" or "reference sequence" refers to any specific known genomic sequence, partial or complete, of any organism or virus that can be used to refer to specific sequences obtained from a subject. For example, the reference genome used for human subjects, as well as for many other organisms, can be found at the National Center for Biotechnology Information at ncbi.nlm.nih.gov. The term "genome" refers to the complete genetic information of an organism or virus expressed in nucleotide sequences.

[0095] In various embodiments, the reference sequence is significantly larger than the reads that align to it. For example, it can be at least about 100 times larger, or at least about 1,000 times larger, or at least about 10,000 times larger, or at least about 10 ⁵ times larger, or at least about 10 ⁶ times larger, or at least about 10 ⁷ times larger.

[0096] In one example, the base sequence is a full-length human genome sequence. Such sequences may be referred to as reference genomic sequences. In another example, the base sequence is limited to a specific human chromosome, such as chromosome 13. In some embodiments, the reference Y chromosome is a Y chromosome sequence from the hg19 version of the human genome. Such sequences may be referred to as reference chromosomal sequences. Other examples of reference sequences include genomes of other species, as well as chromosomes, subchromosomal regions (such as chromatids) of any species.

[0097] In some embodiments, the reference sequence required to perform the alignment may have a length greater than the read length by about 1 to about 100 times. In such embodiments, the alignment and sequencing is considered a targeted alignment or sequencing, rather than a whole genome alignment or sequencing. In these embodiments, the reference sequence typically includes the gene of interest and/or the repeat sequence.

[0098] In various embodiments, the reference sequence is a consensus sequence or other combination obtained from a plurality of individuals. However, in some applications, the reference sequence may be selected from a specific individual.

[0099] The term "clinically significant sequence" as used herein refers to a nucleotide sequence that is known or suspected to be associated with a disease, including a hereditary disease. Determining the absence or presence of a clinically significant sequence may be useful in determining or confirming a diagnosis of a medical condition or in predicting the progression of a disease.

[00100] The term "obtained" as used herein in the context of a nucleic acid or a mixture of nucleic acids means that the nucleic acid(s) is(are) obtained from the source from which it is(are) derived. For example, in one embodiment, a mixture of nucleic acids obtained from two different genomes means that the nucleic acids, such as cDNA, were naturally released by cells through natural processes such as necrosis or apoptosis. In another embodiment, the term "mixture of nucleic acids obtained from two different genomes" means that the nucleic acids were extracted from two different types of cells in a patient.

[00101] The term "based" when used in the context of obtaining a specific quantitative value herein refers to the use of another quantity as an input to calculate a specific quantitative value as an output.

[00102] As used herein, the term "patient sample" refers to a biological sample collected from a patient, i.e., a person receiving medical care or treatment. The patient sample may be any of the samples described herein. In some embodiments, the patient sample is collected using non-invasive procedures, and may be, for example, a peripheral blood sample or a fecal sample. The methods described herein may be applicable to more than just humans. Thus, a variety of veterinary applications are contemplated, in which case the patient sample may be a sample collected from a non-human mammal (e.g., a feline, a pig, an equine, a bovine, etc.).

[00103] The term "biological fluid" as used herein refers to a fluid taken from a biological source and includes, for example, blood, serous exudate, plasma, sputum, lavage, cerebrospinal fluid, urine, semen, sweat, tears, saliva, etc. The terms "blood", "plasma", and "serous exudate" as used herein refer to fractions or processed portions thereof. Similarly, if a sample is obtained by performing a biopsy, taking a wash, smear, etc., the term "sample" refers to a processed fraction or portion obtained by performing a biopsy, taking a wash, smear, etc.

[00104] The term "corresponding" as used herein sometimes refers to a nucleotide sequence, such as a gene or chromosome, that is present in the genome of different subjects and that does not necessarily have the same sequence in all genomes, but serves to define, rather than provide genetic information about, the sequence, such as a gene or chromosome, being examined.

[00105] The term "chromosome" as used herein refers to a hereditary gene carrier of a living cell, which consists of chromatin chains containing DNA and protein components (particularly histones). The standard internationally recognized numbering system for individual chromosomes of the human genome is used herein.

[00106] The term "polynucleotide length" as used herein refers to the absolute number of monomeric subunits (nucleotides) of nucleic acid in a sequence or region of a reference genome. The term "chromosome length" refers to a known chromosome length defined by nucleotide base pairs, such as that given in the NCBI36/hg18 human chromosome assembly, data for which can be found on the Internet at the following address: |genome|.|ucsc|.|edu/cgi-bin/hgTracks?hgsid=167155613&chromInfoPage=.

[00107] The terms "subject" and "patient" as used herein refer to a human subject as well as a non-human subject such as a mammal, an invertebrate, a vertebrate, a fungus, a yeast, a bacterium, and a virus. Although the examples and terminology used herein refer to the human body, the concepts described herein are applicable to the genomes of any plant or animal and may be applied in the fields of veterinary medicine, animal science, laboratory research, etc.

[00108] As used herein, the term "primer" refers to an isolated oligonucleotide that is capable of acting as an initiation point for synthesis when placed under conditions that are inductive with respect to the synthesis of an extension product (e.g., such conditions include the presence of nucleotides, an inducing agent such as DNA polymerase, a suitable temperature and pH). The primer may preferably be single-stranded to ensure maximum amplification efficiency, but may also be double-stranded. If the primer is double-stranded, it is first treated to separate the strands before it is used to obtain extension products. The primer may be an oligodeoxyribonucleotide. The primer is long enough to initiate the process of synthesis of extension products in the presence of an inducing agent. The exact lengths of primers depend on many factors, including temperature, sources of primers, and the methods and parameters used in creating the primers.

5.2. Introduction

[00109] Short tandem repeats (STRs) are found throughout the human genome. Although the biology of STRs is not yet fully understood, emerging evidence suggests that STRs play important roles in fundamental cellular processes.

[00110] Repeat expansions are a special class of microsatellite and minisatellite variants that include STR polymorphisms. Repeat expansions are also called dynamic mutations due to their instability, where short tandem repeats expand beyond a certain size. STR expansions are the underlying cause of more than 20 severe neurological disorders, including amyotrophic lateral sclerosis, Friedreich's ataxia (FRDA), Huntington's disease (HD), and fragile X syndrome. Table 1 presents a small number of pathogenic repeat expansions that differ from repeat sequences in normal samples. The columns illustrate the genes associated with the repeat sequences, the nucleotide sequences of the repeat units, the repeat unit repeat numbers in normal and pathogenic sequences, and the diseases associated with the repeat expansions.

Table 1. Examples of pathogenic repeat expansions Gene Repeat Normal Pathogenic Disease FMR1 CGG 6-60 200-900 Fragile X chromosome AR CAG 9-36 38-62 Spinobulbar muscular atrophy GHT CAG 11-34 40-121 Huntington's disease FXN GAA 6-32 200-1700 Friedreich's ataxia ATXN1 CAG 6-39 40-82 Spinocerebellar ataxia ATXN10 ATTCT 10-20 500-4500 Spinocerebellar ataxia ATXN2 CAG 15-24 32-200 Spinocerebellar ataxia ATXN3 CAG 13-36 61-84 Spinocerebellar ataxia ATXN7 CAG 4-35 37-306 Spinocerebellar ataxia C9ofr72 GGGGCC < 30 100's ABS

[00111] Genetic diseases associated with repeat expansions are heterogeneous in many respects. The size of the repeat unit, the extent of the expansion, the location relative to the affected gene, and the pathogenic mechanism may vary depending on the disease. For example, ABS involves an expansion of the GGGGCC hexanucleotide repeat in the C9orf72 gene, located on the short arm of open reading frame 72 of chromosome 9. In contrast, fragile X syndrome is associated with an expansion of the CGG trinucleotide repeat (triplet repeat) affecting gene 1 ( FMR1) on the X chromosome, which is the cause of mental retardation in fragile X syndrome. The expansion of CGG repeats can result in the failure to express FMRP, the absence of which causes mental retardation in fragile X syndrome and whose presence is essential for normal neurodevelopment. Depending on the length of the CGG repeat, the allele can be classified as normal (not affected by the syndrome), premutation (at risk of developing fragile X disorders), or fully mutated (usually affected by the syndrome). According to various estimates, there are between 230 and 4,000 CGG repeats in the mutated FMR1 genes that cause fragile X syndrome in affected individuals, compared with 60 to 230 repeats in ataxia-prone carriers and 5 to 54 repeats in healthy subjects. FMR1 gene repeat expansion is a cause of autism spectrum disorder, as it has been found that about 5% of patients with autism have FMR1 gene repeat expansion as a cause of autism. McLennan, et al. (2011), Fragile × Syndrome, Current Genomics 12 (3): 216–224. Definitive diagnosis of fragile X syndrome involves genetic testing to determine the number of CGG repeats.

[00112] Many studies have identified various common properties of repeat expansion diseases. Repeat expansion or dynamic mutation typically manifests itself as an increase in the number of repeats, with the rate of such mutation being related to the number of repeats. Rare events, such as loss of a repeat break, can lead to an increased likelihood of allelic expansion; such events are called adverse events. There may be a relationship between the number of repeats in a repeat sequence and the severity of the repeat expansion disease and/or its progression.

[00113] Thus, the detection of repeat expansions plays an important role in the diagnosis and treatment of various diseases. However, determining repeat sequences, especially using reads that do not completely cover the repeat sequence, is associated with several challenges. First, it is difficult to align repeats to a reference sequence because there is no clear one-to-one mapping between the read and the reference genome. Furthermore, even if a read is aligned to a reference sequence, the reads are often too short to completely cover the medically significant repeat sequence. For example, reads can be approximately 100 bp in length. In comparison, repeat expansions can span hundreds to thousands of base pairs. For example, in fragile X syndrome, the FMR1 gene can have over 1,000 repeats spanning over 3,000 bp. Thus, a 100 bp long read cannot match the full length of the repeat expansion. In addition, assembling short reads into a longer sequence does not overcome the problem of short reads and long repeats, since assembling short reads into a longer sequence is difficult due to the ambiguity of the alignment of repeats within one read with repeats in another read.

[00114] Alignment is the primary cause of information loss, either due to incomplete reference sequences, non-unique correspondence between reads and sites on the reference sequence, or significant deviations from the reference sequence. Systematic sequencing errors and other issues that affect read accuracy are secondary causes of failure to detect repeat sequences. In some experimental protocols, approximately 7% of reads are unaligned or have a MAPQ score of 0. Even as researchers work to improve sequencing technology and analysis tools, there will always be a significant number of unaligned and poorly aligned reads. Embodiments of the repeat expansion detection methods described herein rely on detecting unaligned or poorly aligned reads.

[00115] Methods that use long reads to detect repeat expansions have certain disadvantages. In the context of next-generation sequencing, currently available technologies that use longer reads are slower and more prone to errors than technologies that use shorter reads. Moreover, in some applications, long reads are not feasible, such as in cell-free DNA sequencing. Cell-free DNA obtained from maternal blood can be used for prenatal genetic diagnosis. Cell-free DNA exists in fragments that are typically shorter than 200 when using cell-free DNA. Embodiments of the methods described herein use short reads to detect medically significant repeat expansions.

[00116] Moreover, conventional methods are not designed to handle complex loci containing multiple repeats. Important examples of such loci include the HD-causing CAG repeat flanked by a CCG repeat, the FRDA-causing GAA repeat flanked by an adenosine homopolymer, and the spinocerebellar ataxia type 8 (SCA8)-causing CAG repeat flanked by an ACT repeat. An even more striking example is the CCTG repeat in the CNBP gene, expansions of which cause myotonic dystrophy type 2 (DM2). This repeat is adjacent to polymorphic TG and TCTG repeats (JE Lee and Cooper 2009), making precise alignment of reads to this locus particularly challenging. Another type of complex repeat is the polyalanine repeat, which has been linked to at least nine disorders to date (Shoubridge and Gecz 2012). Polyalanine repeats consist of repeats of the a-amino acid codons GCA, GCC, GCG, or GCT.

[00117] Variant clusters can impact genotyping alignment and accuracy (Lincoln et al. 2019). Variants adjacent to low-complexity polymorphic sequences may present additional challenges, as variant detection methods may infer clusters of misrepresented or false variant calls in such genomic regions. This is particularly due to the increased error rates in such regions of sequencing data (Benjamini and Speed 2012; Dolzhenko et al. 2017). One example is the single nucleotide variant (SNV) adjacent to the adenosine homopolymer in the MSH2 gene, which causes Lynch syndrome I (Frogatt et al. 1999).

[00118] Embodiments described herein may handle complex loci as described above. They use a sequence graph as a general flexible model of each target locus.

[00119] In some embodiments, the disclosed methods address the above-mentioned problems in detecting and recognizing repeat expansions by using paired-end sequencing. Paired-end sequencing involves fragmenting DNA to form sequences called inserts. In some protocols, including those used by Illumina, reads from shorter inserts (e.g., tens to hundreds of bp) are referred to as short insert paired-end reads or paired-end reads. In contrast, reads from longer inserts (e.g., on the order of several thousand bp) are referred to as mate-paired reads. As noted above, various embodiments of the methods described herein may use both short insert paired-end reads and long insert mate-paired reads.

[00120] Fig. 1A is a schematic illustration demonstrating certain difficulties in aligning sequence reads to a repeat sequence on a reference sequence, particularly when aligning sequence reads obtained from a sample of a long repeat sequence having a repeat expansion. The lower portion of Fig. 1A illustrates a reference sequence 101 with a relatively short repeat sequence 103, depicted by vertical dashed lines. The middle portion of the figure illustrates a hypothetical sequence 105 of a sample taken from a patient having a long repeat sequence 107 with a repeat expansion, also depicted by vertical dashed lines. The upper portion of the figure shows sequence reads 109 and 111 plotted at the locations of the corresponding sites in the sample 105 sequence. In some of these sequence reads, such as reads 111, some base pairs originate from the long repeat sequence 107, depicted by the vertical dashed lines and circled. Reads 111 having these repeats are potentially difficult to align with the reference sequence 101 because such repeats do not have clear locations on the reference sequence 101. Because these potentially unaligned reads cannot be clearly linked to the repeat sequence 103 in the reference sequence 101, obtaining repeat sequence information and repeat sequence expansion from these potentially unaligned reads 111 is difficult. In addition, since these reads are generally shorter than the long 107 repeat sequence having a repeat expansion, they cannot provide comprehensive information about the type or location of the 107 repeat sequence. In addition, the repeats in the 111 reads make their assembly difficult due to the ambiguity of their locations on the 101 reference sequence and the ambiguity of the relationships between the 111 reads. Reads partially belonging to the long 107 repeat sequence in the sample, half shaded and half black in the illustration, may be aligned with bases that do not belong to the 107 repeat sequence. If the reads contain too few base pairs outside the 107 repeat sequence, the reads may align poorly or not at all. Thus, some of these reads with partial repeats can be analyzed as base reads, while others can be analyzed as anchored reads, as further described below.

[00121] Fig. 1B is a schematic diagram illustrating how paired end reads can be used in some of the described embodiments to overcome the difficulties illustrated in Fig. 1A. In paired end sequencing, sequencing occurs from both ends of nucleic acid fragments in a sample of interest. The bottom portion of Fig. 1B shows a reference sequence 101 and a sample sequence 105, as well as reads 109 and 111 equivalent to those illustrated in Fig. 1A. The top portion of Fig. 1B shows fragment 125 derived from the sequence of sample of interest 105, and the region of primers 131 of read 1, as well as the region of primers 133 of read 2, to produce two paired end reads 135 and 137. Fragment 125 is also referred to as an insert for the paired end reads. In some embodiments, inserts can be amplified with or without PCR. Some repeat sequences, such as those containing a large number of GC or GCC repeats, cannot be sequenced with sufficient quality by traditional methods involving PCR amplification. Such sequences can be amplified without PCR. Other sequences can be amplified using PCR.

[00122] Insert 125 shown in Fig. 1B is obtained from or corresponds to a region of the sequence of sample 105, to the side of which are two vertical arrows shown in the lower half of the figure. In particular, insert 125 comprises a repeating section 127 corresponding to a portion of the long repeat 107 in the sequence of sample 105. The length of the inserts can be adjusted depending on the application. In some embodiments, the inserts can be somewhat shorter than the repeat sequence under study or the repeat sequence having a repeat expansion. In other embodiments, the inserts can have a length similar to the repeat sequence or the repeat sequence with a repeat expansion. In other embodiments, the inserts can even be somewhat longer than the repeat sequence or the repeat sequence with a repeat expansion. In some embodiments, such inserts can be long inserts for performing mate pair sequencing in some embodiments, described further below. Typically, reads derived from inserts are shorter than the repeat sequence. Since inserts are longer than reads, paired end reads can better capture signals over a longer stretch of repeat sequence in a sample than single end reads.

[00123] The illustrated insert 125 has two readout primer regions 131 and 133 at the two ends of the insert. In some embodiments, the readout primer regions are inherent to the insert. In other embodiments, the primer regions are introduced into the insert by ligation or extension. At the left end of the insert, a region of read 1 primers 131 is depicted that allows read 1 primer 132 to hybridize with insert 125. As a result of primer extension 132, reads 1 generate a first read or read 1, designated as 135. At the right end of insert 125, a region of read 2 primers 133 is depicted that allows read 2 primer 134 to hybridize with insert 125, initiating a second read or read 2, labeled as 137. In some embodiments, insert 125 may also include index regions provided with a barcode (not illustrated in the figure), which allows for detection of different samples during multi-channel sequencing. In some embodiments, paired end reads 135 and 137 can be obtained by Illumina sequencing using synthesis platforms. An example of a sequencing workflow implemented on such a platform is further described below in the Sequencing Methods section, which generates two paired-end reads and two index reads.

[00124] The paired end reads obtained by the method depicted in Fig. 1B can then be aligned to a reference sequence 101 having a relatively short repeat sequence 103. In this way, the relative location and direction of the pair of reads is known. This allows an unaligned or poorly aligned read, such as that indicated in circle 111, to be indirectly linked to the sequence of the relatively long repeats 107 in the sample sequence 105 via a corresponding paired read 109, as indicated in the lower portion of Fig. 1B. In an illustrative example, the reads obtained from paired end sequencing are about 100 bp and the inserts are about 500 bp. In this illustrative configuration, the relative locations of the two paired end reads are about 300 bp from their 3' ends and are in opposite directions. The relationship between pairs of reads allows for better association of reads with repeat regions. In some cases, the first read in a pair aligns with a non-repeat sequence flanking the repeat region on the reference sequence, while the second read in the pair does not align with the reference properly. See, for example, the information on the pair of reads 109a and 111a shown in the lower half of Fig. 1B, where the first read is the left read 109a and the second read is the right read 111a. Given the pairing of the two readings 109a and 111a, the second reading 111a can be associated with the region of repeats 107 in the sequence of the sample 105, despite the fact that the second reading 111a cannot be aligned with the reference sequence 101. Knowing the distance and direction of the second reading 111a relative to the first reading 109a, it is possible to further determine the location of the second reading 111a within the long region of repeats 107. In case there is a gap between the repeats in the second reading 111a, it is also possible to determine the location of the gap relative to the reference sequence 101. In the present description, a read such as the left reading 109a, which is aligned with the reference, is called a base read. A read such as the right reading 111a, which is not aligned with the reference sequence, but is paired with the base read, is called an anchored read. Thus, the unaligned sequence may be associated with a repeat expansion. Thus, short reads can be used to detect long repeat expansions. Although the problem of detecting repeat expansions usually increases with the length of the expansion due to the difficulty of performing sequencing, the methods described herein can detect a higher signal from longer repeat expansion sequences than from shorter repeat expansion sequences. This is because as the length of the repeat sequence or repeat expansion increases, more reads will be anchored in the expansion region, and more reads will have a probability of falling in the repeat region, and more repeats can occur with each read.

[00125] In some embodiments, the described methods include analyzing the frequency distribution of the number of repeats detected within the base and anchored reads. In some embodiments, only the anchored reads are analyzed. In other embodiments, both the base and anchored reads are analyzed. The distribution of the sample under study can be compared to a criterion obtained empirically or theoretically and used to separate unaffected samples from affected samples. In this way, it can be determined whether the sample under study contains the repeat expansion in question and, accordingly, recognition can be performed.

[00126] The methods and apparatus described herein may utilize next generation sequencing (NGS) technology that enables massively parallel sequencing. In certain embodiments, clonally amplified DNA templates or single DNA molecules are sequenced within a flow cell by massively parallel sequencing (e.g., as described in Volkerding et al. Clin Chem 55:641-658 [2009]; Metzker M Nature Rev 11:31-46 [2010]). NGS sequencing technologies include, but are not limited to, pyrosequencing, reverse dye terminator synthesis sequencing, oligonucleotide probe ligation sequencing, and ion semiconductor sequencing. DNA from individual samples can be sequenced individually (i.e., single-channel sequencing) or DNA from multiple samples can be pooled and sequenced as indexed genomic molecules (i.e., multi-channel sequencing) in a single sequencing run to generate up to several hundred million DNA sequence reads. Examples of sequencing technologies that can be used to obtain sequence information in the present method are further described below.

[00127] Various repeat expansion assays using DNA samples involve aligning or matching sequence reads from a sequencer to a reference sequence. The reference sequence may be a whole genome sequence, a chromosome sequence, a subchromosomal region sequence, etc. From a computational standpoint, repeats create ambiguities in alignment, which in turn may lead to errors at the level of whole chromosome counting. In various embodiments, paired end reads in combination with adjustable insert length may help to resolve ambiguities in repeat sequence alignment and detect repeat expansion.

Genotyping variants at the repeat sequence locus using sequence graph

[00128] Fig. 1C is a flow chart illustrating a process 140 for genotyping a genomic locus comprising a repeat sequence, according to some embodiments. In some embodiments, the genetic locus is predetermined in a catalog of variants comprising genomic positions and a structure of loci at the genomic positions. Figs. 1D, 1E, and 1F illustrate three different sequence graphs according to some embodiments.

[00129] Fig. 1G illustrates a schematic diagram of a process for determining genotypes of variants of an HTT locus containing two CTP sequences, according to some embodiments. Panel (a) of Fig. 1G depicts a portion of a variant catalog containing genomic loci and locus specifications, in particular their structure. For example, excluding repeats, the sequence at the HTT locus is CAGCAACAGCGG (SEQ ID NO: 2); the sequence at the CNBP locus is CAGGCAGACA (SEQ ID NO: 3).

[00130] Fig. 1H illustrates a schematic diagram of the process for determining genotypes of variants at a Lynch I locus containing an OHB and a CTP, according to some embodiments. In Fig. 1H, box 162 illustrates the general structure of locus specifications, and box 163 depicts a specific example of a Lynch I (MSH2) locus specification.

[00131] In the catalog of variants, the locus structure is specified using a limited subset of the regular expression syntax. For example, the HD-associated repeat region may be defined by the expression (CAG)* CAACAG(CGG)* or SEQ ID NO: 2 (excluding repeats), meaning that it contains variable numbers of CAG and CCG repeats separated by a CAACAG gap; the region associated with the FRDA region corresponds to the expression (A)*(GAA)*; the region associated with SCA8 corresponds to the expression (CTA)*(CTG)*; the DM2 repeat region, consisting of three contiguous repeats, is defined by the expression (CAGG)*(CAGA)*(CA)* or SEQ ID NO: 3 (excluding repeats); the MSH2 OHB is contiguous with the homopolymer causing Lynch syndrome I and corresponds to the expression (A|T)(A)*.

[00132] In addition, regular expressions may contain multiallelic or "degenerate" base symbols, which may be specified using the International Union of Pure and Applied Chemistry (IUPAC) notation ("List of partially determined bases in nucleotide sequences. Recommendations", 1984. Nomenclature Committee of the International Union of Biochemists (IUBN)" 1986).

[00133] Partially defined bases corresponding to bases in degenerate codons are referred to herein as degenerate bases. Degenerate bases allow certain classes of imperfect DNA repeats to be represented, in which, for example, different bases can occur at the same position. Using this notation, polyalanine repeats may be encoded by the expression (GCN)*, and polyglutamine repeats may be encoded by the expression (CAR)*.

[00134] In some embodiments, the repeat sequence included in the genomic locus includes a short tandem repeat (STR) sequence. In some embodiments, the FTR expansion is associated with fragile X syndrome, amyotrophic lateral sclerosis (ALS), Huntington's disease, Friedreich's ataxia, spinocerebellar ataxia, spinobulbar muscular atrophy, myotonic dystrophy, Machado-Joseph disease, or dentato-rubro-pallido-Lewis atrophy.

[00135] Process 140 includes collecting nucleic acid sequence reads of a sample of interest from a database. See block 142. In some embodiments, the nucleotide sequence reads are initially aligned to a reference genome, but in this case, the process re-aligns the sequence reads to the genomic locus of interest, as described below. In alternative embodiments, the reads may be directly aligned to a sequence graph without first being aligned to a reference genome.

[00136] Process 140 aligns sequence reads with a genomic locus sequence that includes one or more repeat sequences. See block 144. The genomic locus sequence is represented by data stored in system memory, which also stores a sequence graph data structure. The sequence graph includes a directed graph where nodes represent nucleotide sequences and directed edges connect nodes. The nucleotide sequence represented by a node includes one or more nucleotide bases. The sequence graph includes one or more proper simple cycles. Each proper simple cycle is a repeat sequence of one or more repeat sequences. Each repeat sequence includes repeats of a repeat unit of one or more nucleotides.

[00137] In some embodiments, sequence reads are initially aligned to a reference genome to determine genomic coordinates of the reads before a subset of the initially aligned reads are aligned to one or more sequence graphs representing one or more sequences of interest. In some embodiments, the initially aligned reads are aligned to sequence graphs to determine repeat expansions in a range of several tens to several thousand regions (each region corresponding to a sequence graph). The total number of initially aligned reads that are re-aligned to sequence graphs during each application of embodiments may range from thousands to many millions of reads.

[00138] In some embodiments, reads initially aligned to a sequence or locus of interest are selected as a subset of reads, the subset is then aligned to repeat sequences, each of which is represented in a sequence graph, wherein the sequence graph has one or more of its own simple cycles, and the natural simple cycles represent one or more repeat sequences. In various embodiments, a read that is within about 10, 50, 100, 500, 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 50,000, 100,000 bases of the sequence or locus of interest is considered to be in the vicinity of the sequence or locus of interest. In some embodiments, a read that is within about 1,000, 2,000, 3,000, 4,000, 5,000, 6,000, 7,000, 8,000, 9,000, or 10,000 bases of a locus of interest is adjacent to the locus of interest. Some of the raw reads may be initially poorly aligned because they include, among other things, repeat sequences that are difficult to align in an unambiguous manner. In some embodiments, reads that are initially poorly aligned (e.g., as measured using an alignment score) but are paired with a read that aligns to the locus of interest or a space near it (in a pair of paired-end reads) are aligned to the sequence graph. In some embodiments, reads initially aligned to off-target regions that represent known hotspots for misaligned reads are aligned to the sequence graph.

[00139] Figs. 1D, 1E, and 1F illustrate three different sequence graphs according to some embodiments. Fig. 1D illustrates a first sequence graph 1100 representing a first genomic locus comprising a repeat sequence having a CAG trinucleotide repeat unit. The first sequence graph 1100 includes vertices 1102 and 1112, respectively representing two flanking sequences. The first sequence graph also includes a vertex 1106 representing a repeat sequence comprising a CAG trinucleotide repeat unit. The first sequence graph includes a directed edge 1104 connecting vertex 1102 (the flanking sequence) and vertex 1106 (the CAG repeat sequence) in a direction from vertex 1102 to vertex 1106. The direction of the edge indicates the relative position of the two nucleotide sequences. The first sequence graph also includes a directed edge 1104 connecting node 1102 (the flanking sequence) and node 1106 (the CAG repeat sequence) in the direction from node 1102 to node 1106. The first sequence graph also includes a directed edge 1110 connecting node 1106 (the CAG repeat sequence) and node 1112 (the flanking sequence) in the direction from node 1106 to node 1112. The first sequence graph also includes a proper simple cycle 1108, which is a repeat sequence containing a CAG repeat unit (illustrated by node 1106) that is repeated one or more times. The path from the starting node to the ending node of the sequence graph is the sequence of the genomic locus, which may contain nucleotides located near the repeat sequence, such as flanking sequences.

[00140] Fig. 1E illustrates a second sequence graph 1200 representing a second genomic locus. The second sequence graph 1200 includes vertices 1202 and 1224, respectively representing two flanking sequences. The second sequence graph also includes vertex 1206 and vertex 1216, representing a repeat sequence comprising a CAG trinucleotide repeat unit and a repeat sequence comprising a CCG trinucleotide repeat unit, respectively. The second sequence graph also includes vertex 1212, representing a CAACAG sequence that does not have repeats. The second sequence graph includes directed edges 1204, 1210, 1214 and 1220. These directed edges directionally connect vertices 1202, 1206, 1212, 1216 and 1224, as shown in the illustration. The second sequence graph also includes a proper simple cycle 1208, which is a repeat sequence containing a repeating unit CAG (illustrated by vertex 1206), which is repeated one or more times. The second sequence graph also includes a proper simple cycle 1218, which is a repeat sequence containing a repeating unit CCG (illustrated by vertex 1216), which is repeated one or more times.

[00141] Fig. 1F illustrates a third sequence graph 1300 representing a third genomic locus. The third sequence graph 1300 is similar to the second sequence graph 1200, but includes two alternative paths representing two alleles of CAC and CAT. The two alleles may be either SNP or NBC alleles. Directed edge 1310, vertex 1312, and directed edge 1314 represent the first CAC allele. Directed edge 1316, vertex 1318, and directed edge 1320 represent the second CAT allele. The third sequence graph includes elements otherwise similar to the elements of the second sequence graph, including vertices 1302, 1306, 1322, and 1328. It also includes proper simple cycles 1308 and 1324, which point to the CAG and CCG repeat sequences. It additionally includes directed edges 1304 and 1326.

[00142] In some embodiments, sequence reads are aligned to a sequence graph using techniques described below.

[00143] 1. The kmer index is constructed based on the entire graph in such a way that, given a kmer value from a sequence, all nodes of the graph at which such kmer begins or ends can be recalculated. In some cases, a kmer may begin at one node and end at another node.

[00144] 2. Two subgraphs are registered relative to each graph point: one in the forward direction of the kmer and one in the reverse direction. The subgraphs are subject to "unfolding" of the repeat expansion to the remaining length of the read, while they do not have any nodes located further from the kmer hit point compared to the remaining length of the read, and it is assumed that no repeat expansion occurs. The procedure is a "breadth-first" search method and is used to generate a data structure containing the following elements:

[00145] - Sequence of all node sequences (including extended repeats) on a subgraph

[00146] - A node index that allows a node identifier to be easily obtained from an offset in the sequence when performing Smith-Waterman backtracking.

[00147] - Sequence of node end offsets with incoming edges for each node start offset

[00148] - An index for each node, allowing easy determination of whether the base is at the start of the node or not, and also counting all end offsets of previous nodes.

[00149] 3. Alignment serves to:

[00150] - Maintaining affine gaps.

[00151] - Find the best sequence alignment(s) given the above information and the penalty matrix.

[00152] Two different interfaces are available:

[00153] - The best alignment and second best alignment score are reported.

[00154] - The entire array of best alignments, as well as the second best alignment score.

[00155] The alignments are general alignments that define a penalty for the gap between a potential kmer and the start of the aligned sequence. In some implementations, compile-time parameters are adjusted.

[00156] The matrix filling algorithm used is available in two implementation variants:

[00157] - Sequential cycles with complexity N*M.

[00158] - Fixed-size sequential loops with a fixed-length compile-time option defaulting to 16, which gcc automatically recognizes and translates to SSE or AVX vector instructions on the CPU.

[00159] In some embodiments, a repeating unit of one or more repeat sequences comprises at least one partially defined nucleotide. In some embodiments, a particular repeating unit comprises degenerate codons.

[00160] In some embodiments, the one or more native simple cycles comprise two or more native simple cycles representing two or more repeat sequences. See, for example, Fig. 1E, Fig. 1F, and Fig. 1G, panel (b).

[00161] In some embodiments, the sequence graph further comprises two or more alternative pathways for two or more alleles. See, for example, Fig. 1F, reference numerals 1312 and 1318. See also Fig. 1H, reference numerals 165 and 167a, indicating the Lynch I (MSH2) locus, where the upper pathway includes a vertex for nucleic acid base A and the lower pathway includes a vertex for nucleic acid base T.

[00162] In some embodiments, two or more alleles comprise a deletion or a substitution. In some embodiments, the substitution comprises a single nucleotide variant (SNV) or a single nucleotide polymorphism (SNP). See, for example, Fig. 1F, reference numbers 1312 and 1318.

[00163] In some embodiments, aligning a sequence read with a sequence graph includes: finding a kmer correspondence between the sequence read and a path of the sequence graph and then expanding the path to a complete alignment. In some embodiments, the alignment includes extracting a subgraph of the path; "unrolling" cycles in the subgraph to obtain a directed acyclic graph; and performing a Smith-Waterman alignment on the sequence read over the directed acyclic graph.

[00164] In some embodiments, aligning a sequence read to a sequence graph includes pruning the graph by removing low-confidence ends of the alignments. After aligning the read to the graph, the method searches for other similar alternative alignments. This is accomplished by re-aligning the initial read to the paths of the graph that overlap the path of the initial alignment. This allows one or both ends of the initial alignment to be determined to have low confidence; such low confidence indicates that they could have been aligned in a different way. The ability to detect high and low confidence portions of the alignment allows one to accurately determine which genetic variants are supported within the read.

[00165] In some embodiments, aligning a sequence read to a sequence graph includes combining alignments by: aligning subsequences of the read to the sequence graph; and combining the subsequence alignments to fully align the sequence read.

[00166] In some embodiments, the process also generates a sequence graph based on a locus specification that includes the structure of the genomic locus. In some embodiments, the locus specification is defined in a catalog of variants, as explained above.

[00167] See also panels (b)-(d) of Fig. 1G for schematic illustrations of read alignment to the HTT locus sequence graph. Fig. 1H schematically depicts locus analyzers 164 for performing read alignment to the sequence graph, including relative to the Lynch I locus (165).

[00168] Method 140 further includes determining one or more genotypes of one or more repeat sequences using sequence reads aligned to a sequence graph. See block 140. See also Panel (e) of Fig. 1G depicts the determination of two CTPs (CAG and CCG) at the HTT locus. The sequence on the left, including the CAG repeats, is CAGCAGCAGCAGCAG (SEQ ID NO: 4). The sequence on the left, including the CCG repeats, is CCGCCGCCGCCGCCG (SEQ ID NO: 5).

[00169] Fig. 1H depicts a Variant Genotyper software module (168) for calling variants at the Lynch I locus, including OHBs with A/T alleles (169a) and an A repeat monomer (169b). Fig. 1H also shows Variant Analyzer software modules (166) for managing sequence alignment data and feeding it to Variant Genotyper software (168), as well as implementation paths for Variant Analyzer software for working with OHBs with A/T alleles (167a) and an A repeat monomer (167b). The locus results obtained using Genotyper software are illustrated in Fig. 1H, box 170, which also shows the genotype of OHBs with A/T alleles (171a) and an A repeat monomer (171b).

[00170] In some embodiments, the sequence graph includes two alternative pathways for two alleles, and the method further includes genotyping the two or more alleles using sequence reads aligned with the two or more alternative pathways. In some embodiments, genotyping the two or more alleles includes covering the two or more alternative pathways to a probability model for determining probabilities of the two or more alleles. In some embodiments, the probability model models the probability of an allele as a function of the allele coverage, wherein the function is selected from a Poisson distribution, a negative binomial distribution, a binomial distribution, or a beta-binomial distribution.

[00171] In some embodiments, the probability function is a Poisson distribution and its rate parameter is estimated from the read length and average depth observed at the genomic locus.

[00172] In the Poisson model, the allele probability is expressed as follows:

[00173] P(Y=y) = (C ^y × e ^-C )/y!

[00174] • y represents the coverage of base reads

[00175] • C is the average depth observed at a genomic locus

[00176] In some embodiments, the average depth C is determined as follows.

[00177] C=LN/G

[00178] • G is the length of the genomic locus

[00179] • L represents the read length

[00180] • N represents the total number of reads

GraphTools Library

[00181] In some embodiments, the GraphTools library is used when working with sequence graphs. The library contains key schematic representations of graphs (the graphs themselves, graph paths, and graph alignments), operations with them, and algorithms for aligning linear sequences with graphs.

[00182] In some embodiments, a sequence graph consists of nodes and directed edges. Graphs may include their own simple cycles (edges connecting a node to itself), but not include other cycles. Nodes contain sequences consisting of basic bases and degenerate base codes defined by IUPAC.

[00183] A path in a graph is defined by a sequence of nodes through which the path passes, as well as the starting point of the path at the first node and the ending point at the last node. Positions are specified using a zero-based "half-open" coordinate system. The library provides notations for a variety of path operations, including path expansion and contraction, overlap tests, and path merging.

[00184] Graph alignments comprise encodings of how linear query sequences (typically read sequences) are aligned to graphs. In some embodiments, a graph alignment comprises a graph path and a sequence of linear alignments that define an alignment of the query sequence to nodes of the graph path. Using appropriate path operations, graph alignments can be pruned or combined with other graph alignments. Path pruning provides a mechanism for eliminating low-confidence alignment ends, while alignment merging is used by graph alignment algorithms to combine a full alignment of a query sequence with subsequence alignments (e.g., kmer). In some embodiments, an alignment algorithm operates by finding a kmer correspondence between a query sequence and a graph and then expanding that correspondence to a full alignment. In some embodiments, the alignment includes extracting a subgraph along the path that matches the kmer correspondence (unrolling cycles in the process). Then, a Smith-Waterman alignment is performed with respect to the resulting directed acyclic graph. In some implementations, the algorithm supports the definition of penalties for affine gaps and is written using constant-length cycles for SIMD compilers to generate code.

[00185] In some embodiments, a path in a graph may be obtained using a search algorithm that includes expanding or contracting the path by increasing or decreasing the number of repetitions of a repeating link represented by its own simple cycle until the alignment reaches a search criterion or convergence (e.g., when the alignment score reaches a maximum value).

[00186] In some embodiments, based on a sequence graph, a plurality of paths in the graph are generated, wherein each path in the graph represents a specific number of repetitions of a repeating unit represented by its own simple cycle. The requested sequence is aligned with the plurality of paths in the graph, and then the path that meets the alignment criterion is selected to perform the graph alignment.

Structure of the method of application

[00187] Some embodiments are designed as a general method for performing targeted genotyping of variants (Figure 1H). During each cycle, the program attempts to genotype a set of variants,

[00188] variants described in the catalog file. Variants located in close proximity to each other are grouped into the same locus. The locus structure is specified using a limited subset of regular expression (RE) syntax. REs contain alphabetically ordered sequences of basic base symbols and degenerate base codes as defined by IUPAC; they must contain one or more of the following expressions (<sequence>)?, (<sequence a>|<sequence b>), (<sequence>)*, (<sequence>)+, possibly separated by sequence breaks. These expressions correspond to insertions/deletions, substitutions, zero or more repetitions of a sequence, and at least one repetition of a sequence, respectively. In addition, each locus description contains a set of core regions for that locus, as well as the core coordinates of each constituent variant.

[00189] The bulk of the work is organized around objects of the LocusAnalyzer class, which synthesizes a sequence graph representing the locus of the corresponding RE during the initialization process. After initialization, the locus analyzer processes the corresponding reads by aligning them to the graph and then passing the resulting alignments to VariantAnalyzer software, which is designed to work with all variants contained in the locus. Variant Analyzer software collects information relevant to the genotyping process of the associated variant and passes it to Genotyper software, which is responsible for performing the genotyping process. The results obtained by each unit of Genotyper software are then used to generate an output file in VCF format.

[00190] For example, the LocusAnalyzer, which is responsible for processing the locus for the pathogenic variant associated with Lynch syndrome I, uses an ONV analyzer and a CTP analyzer (Figure S1, right panel).

5.8. Using Genotyper software to work with deletions

[00191] Some CTPs may have a small insertion or deletion nearby. Such deletions are modeled as additional subgraphs of the flanking sequences of the CTP. The number of reads mapped to each allele (or pathway in the graph) is modeled using a Poisson distribution whose rate parameter is estimated from the average depth and length of reads observed at the locus. The genotype probability is calculated using a Bayesian framework.

5.9 Identification of repeat expansions

[00192] Using the embodiments described herein, various inherited diseases related to repeat expansions can be identified with high efficiency, sensitivity, and/or selectivity compared to traditional methods. In some embodiments, the present invention provides methods for identifying and recognizing medically significant repeat expansions, such as the CGG repeat expansion that causes mental retardation in fragile X syndrome, using sequence reads that do not completely span the repeat sequence. Short reads, such as 100 bp reads, are not long enough to perform sequencing of multiple repeat expansions. However, when analyzed using the methods described, repeat expansion samples exhibit a statistically significant excess of reads containing a large amount of repeat sequence. In addition, extremely large repeat expansions contain misaligned read pairs, where both reads consist entirely or almost entirely of repeat sequence. Normal samples are used to determine background expectations.

[00193] The conventional wisdom is that repeat expansions cannot be detected without using a read that spans the entire repeat. Existing approaches to detecting repeat expansions use targeted sequencing with long reads, and in some cases such reads fail due to insufficient length, making them unable to span the repeat sequence. The results of some of the described embodiments were unexpected, in part because they used normal (non-target) sequence data and a read length of approximately 100 bp, but they provide very high sensitivity in detecting repeat expansions. The methods described herein detect the number of repeat units in a repeat expansion using paired reads having an insert length (i.e., two sequence reads and an intervening sequence) that is shorter than the length of the entire repeat sequence.

[00194] With respect to the details of methods for determining the presence of a repeat expansion in accordance with some embodiments, Fig. 2 is a flow chart depicting a high level of embodiments for determining the presence or absence of a repeat expansion in a repeat sequence in a sample. A repeat sequence is a nucleotide sequence that contains repeated short sequences called repeat units. Table 1 above sets forth examples of repeat units, the number of repeats of repeat units in the repeat sequences relative to normal and pathogenic sequences, genes associated with the repeat sequences, and diseases associated with the repeat expansion. The process 200 illustrated in Fig. 2 begins with obtaining paired-end reads of the sample under study. See block 202. The paired-end reads are processed to align with a reference sequence containing the repeat sequence of study. In some contexts, the alignment process is also referred to as the mapping process. The sample under study contains nucleic acid and may be in the form of a biological fluid, tissue, etc., as further described in the Sample section below. The sequence reads are subjected to an alignment process to match them to a reference sequence. Various tools and alignment algorithms may be used to align the reads to the reference sequence, as described elsewhere herein. Typically, when executing the alignment algorithms, some reads will align successfully to the reference sequence, while other reads may not align as successfully or may align poorly to the reference sequence. Reads that are consistently aligned to the reference sequence are associated with sites on the reference sequence. The aligned reads and their associated sites are also referred to as sequence tags. As explained above, aligning some sequence reads containing a large number of repeats to the reference sequence is more challenging. If a read is aligned with a reference sequence and the mismatch rate is above a certain criterion, the read is considered poorly aligned. In various embodiments, reads are considered poorly aligned if they are aligned with at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 mismatches. In other embodiments, reads are considered poorly aligned if they are aligned with at least about 5% mismatches. In other embodiments, a read is considered poorly aligned if it is aligned with at least about 10%, 15%, or 20% mismatches.

[00195] As shown in Fig. 2, the process 200 describes determining base and anchored reads in paired end reads. See block 204. Base reads are reads between paired end reads that are aligned with the repeat sequence of interest or the space next to it. For example, a base read can be aligned with a location on a reference sequence that is separated from the repeat sequence by a sequence length that is less than the length of the insert sequence. The separation length can be shorter. For example, a base read can be aligned with a location on a reference sequence that is separated from the repeat sequence by a sequence length that is less than the sequence length of the base read or less than the total sequence length of the base read and the sequence that connects the base read to the anchored read (insert length minus anchored read length). In some embodiments, the repeat sequence of interest may be a repeat sequence in the FMR1 gene, including repeats of the CGG repeat unit. In a normal reference sequence, the repeat sequence in the FMR1 gene includes about 6 to 32 repeats of the CGG repeat unit. Since the repeats expand to over 200 copies, there is a potential for pathogenicity of the repeat expansion, which causes fragile X syndrome. In some embodiments, a read is considered to be aligned near the sequence of interest when it is aligned within 1,000 bp of the repeat sequence of interest. In other embodiments, this parameter can be adjusted, for example, within about 100 bp, 200 bp, 300 bp, 400 bp, 500 bp, etc. bp, 600 bp, 700 bp, 800 bp, 900 bp, 1,500 bp, 2,000 bp, 3,000 bp, 5,000 bp, etc. In addition, the process also identifies anchored reads, which are reads that are paired with base reads but are poorly aligned or cannot be aligned to their reference sequence. More information on poorly aligned reads is provided above.

[00196] The process 200 further includes determining the probability of the presence of a repeat expansion of the repeat sequence in the sample under study, based at least in part on the determined anchored reads. See block 206. This step of the determining process may include a variety of suitable analysis and calculation methods, as further described below. In some embodiments, the process uses the determined base reads as well as the anchored reads to determine the probability of the presence of a repeat expansion. In some embodiments, the number of repeats in the determined base and anchored reads is analyzed and compared to one or more criteria derived theoretically or derived from empirical data on the affected control samples.

[00197] In various embodiments described herein, repeats are obtained as in-frame repeats, wherein two repeats of the same repeat unit fall within the same reading frame. A reading frame is a way of dividing a sequence of nucleotides in a nucleic acid molecule (DNA or RNA) into a set of consecutive, non-overlapping triplets. During translation, the triplets encode amino acids, and such triplets are called codons. Thus, any sequence has three possible reading frames. In some embodiments, repeats are counted according to three different reading frames, and the one count of the three that shows the highest value is considered the number of corresponding repeats in the read.

[00198] An example of a process that includes performing an additional step and assays is illustrated in Fig. 3. Fig. 3 is a flow chart illustrating a process 300 required to detect a repeat expansion using paired end reads having a large number of repeats. The method 300 includes additional steps for processing a test sample. The process begins with sequencing a test sample containing nucleic acids to obtain paired end reads. See block 302. In some embodiments, the test sample may be obtained and prepared in a variety of ways, as further described below in the "Samples" section. For example, the test sample may be a biological fluid, such as plasma, or any suitable sample, as described below. The sample may be obtained using a non-invasive procedure, such as a routine blood draw. In some embodiments, the test sample comprises a mixture of nucleic acid molecules, such as csDNA molecules. In some embodiments, the test sample is a maternal plasma sample containing a mixture of fetal and maternal DNA molecules.

[00199] Prior to sequencing, nucleic acids are extracted from the sample. Suitable extraction processes and apparatus are described elsewhere herein. In some embodiments, the apparatus co-processes DNA from multiple samples to multiplex libraries and sequence data. In some embodiments, the apparatus 20 processes DNA from eight or more test samples in parallel. As described below, the sequencing system may process the extracted DNA to produce a library of encoded (e.g., barcoded) DNA fragments.

[00200] In some embodiments, nucleic acids in a test sample may be further processed to prepare libraries for single-channel or multi-channel sequencing, as further described below in the "Preparing Libraries for Sequencing" section. After processing and preparing the samples, nucleic acid sequencing may be performed in a variety of ways. In some embodiments, a variety of next-generation sequencing platforms and protocols may be used, as further described in the "Sequencing Methods" section below.

[00201] Regardless of the particular sequencing platform and protocol, in block 302, at least a portion of the nucleic acids contained in the sample are sequenced to generate tens of thousands, hundreds of thousands, or millions of sequence reads, such as 100 bp reads. In some embodiments, the reads include paired end reads. In other embodiments, including those described below with respect to FIG. 5, in addition to paired end reads, single end long reads spanning more than hundreds of thousands or tens of thousands of bases can be used to determine the sequence of repeats. In some embodiments, the sequence reads comprise about 20 bp, about 25 bp, about 30 bp, about 35 bp, about 36 bp, about 40 bp, about 45 bp, or about 50 bp. o., about 50 bp, about 55 bp, about 60 bp, about 65 bp, about 70 bp, about 75 bp, about 80 bp, about 85 bp, about 90 bp, about 95 bp, about 100 bp, about 110 bp, about 120 bp, about 130 bp, about 140 bp, about 150 bp, about 200 bp, about 250 bp, about 300 bp, about 350 bp o., about 400 bp, about 450 bp, or about 500 bp. When generating paired-end reads, technological advances are expected to enable single-end reads greater than 500 bp and reads greater than 1,000 bp.

[00202] Process 300 describes aligning the results of the paired end reads obtained from block 302 with a reference sequence that includes a repeat sequence. See block 304. In some embodiments, the repeat sequence has a tendency to expand. In some embodiments, the repeat expansion is known to be associated with a genetic disease. In other embodiments, the repeat expansion of the repeat sequence has not previously been studied to establish an association with a genetic disease. The methods described herein can detect a repeat sequence and a repeat expansion independent of any associated pathology. In some embodiments, the reads are aligned with a reference genome; example: hg18. In other embodiments, the reads are aligned with a region of the reference genome, such as a chromosome or a segment of a chromosome. Reads that are uniquely aligned with a reference genome are referred to as sequence tags. In one embodiment, at least about 3 x 10 ⁶ qualified sequence tags, at least about 5 x 10 ⁶ qualified sequence tags, at least about 8 x 10 ⁶ qualified sequence tags, at least about 10 x 10 ⁶ qualified sequence tags, at least about 15 x 10 ⁶ qualified sequence tags, at least about 20 x 10 ⁶ qualified sequence tags, at least about 30 x 10 ⁶ qualified sequence tags, at least about 40 x 10 ⁶ qualified sequence tags, or at least about 50 x 10 ⁶ qualified sequence tags are obtained from reads that map unambiguously to a reference genome.

[00203] In some embodiments, the process may filter the sequence reads prior to alignment. In some embodiments, the read filtering is a quality filtering process performed using sequencer software to filter out erroneous and low-quality reads. For example, Illumina's Sequencing Control Software (SCS) and Consensus Assessment of Sequence and Variation are used to filter out erroneous and low-quality reads by converting raw image data generated by sequencing reactions into intensity values, base calls, quality scores, and additional formats for collecting biologically relevant information for a further analysis step.

[00204] In certain embodiments, reads obtained by a sequencing machine are in an electronic format. The alignment process is performed by a computing device as described below. Individual reads are compared to a reference genome, which is often large (millions of base pairs), to determine points where the reads unambiguously match the reference genome. In some embodiments, alignments allow limited mismatches between reads and the reference genome. In some cases, mismatches of 1, 2, 3, or more corresponding base pairs in the reference genome are allowed and still perform alignment. In some embodiments, reads are considered aligned reads when the reads are aligned to a reference sequence that contains no more than 1, 2, 3, or 4 base pairs. Accordingly, unaligned reads are reads that cannot be aligned or are poorly aligned. Poorly aligned reads are reads with more mismatches than aligned reads. In some embodiments, reads are considered aligned reads when the reads are aligned to a reference sequence that contains no more than 1%, 2%, 3%, 4%, 5%, or 10% of the nucleotide base pairs.

[00205] After aligning the paired-end reads to a reference sequence containing the repeat sequence of interest, process 300 determines base and anchor reads between the paired-end reads. See block 306. As mentioned above, base reads are paired-end reads that are aligned to or near the repeat sequence. In some embodiments, base reads are paired-end reads that are aligned within 1 kb of the repeat sequence. Anchor reads are paired with base reads, but fail to align or are poorly aligned with the reference sequence, as described above.

[00206] The process 300 analyzes the number of repeats of the repeat units in certain base and/or anchored reads to determine the presence or absence of a repeat sequence expansion. More specifically, the process 300 includes using the number of repeats in the reads to obtain high-volume read numbers in the base and/or anchored reads. High-volume reads are reads with more repeats than a threshold. In some embodiments, high-volume reads are obtained only from anchored reads. In other embodiments, high-volume reads are obtained from both base and anchored reads. If, in some embodiments, the number of repeats approaches the maximum possible number of repeats for a read, the read is considered a high-volume read. For example, if a read is 100 bp and the repeat unit in question is 3 bp, i.e., the maximum number of repeats will be 33. In other words, the maximum value is calculated based on the length of the paired end reads and the length of the repeat unit. In particular, the maximum number of repeats can be obtained by dividing the length of the read by the length of the repeat unit and rounding the resulting value down. In this example, various embodiments may determine 100 bp reads having at least about 28, 29, 30, 31, 32, or 33 repeats in the large reads. The number of repeats in the large reads may be adjusted up or down based on empirical factors. In various embodiments, the threshold value for the large reads is at least about 80%, 85%, 90%, or 95% of the maximum number of repeats.

[00207] Next, in the process 300, a possible presence of a repeat expansion in the repeat sequence is determined based on the number of high-volume reads. See block 310. In some embodiments, the analysis compares the obtained high-volume reads with a recognition criterion and makes a determination of the likely presence of a repeat expansion when the criterion values are exceeded. In some embodiments, the recognition criterion is obtained based on a distribution of reads of high-volume control samples. For example, a plurality of control samples that have or are suspected of having a normal repeat sequence are analyzed, and high-volume reads are obtained for the control samples in the same manner as described above. The distribution of high-volume reads can be performed relative to the control samples and the probability of an unaffected sample occurring with high-volume reads exceeding a specific value can be estimated. This probability allows for determining sensitivity and selectivity given a recognition criterion set relative to this specific value. In some embodiments, the recognition criterion is set to a threshold value, and thus the probability of an unaffected sample occurring with high-volume reads exceeding the threshold value is less than 5%. In other words, the p-value is less than 0.05. In these embodiments, as the repeats expand, the repeat sequence becomes longer, and it becomes possible for more reads to appear exclusively within the repeat sequence, and it becomes possible for more reads of the sample to be obtained. In various alternative embodiments, a more conservative discrimination criterion can be selected such that the probability of an unaffected sample with large reads exceeding the threshold is less than about 1%, 0.1%, 0.01%, 0.001%, 0.0001%, etc. It should be understood that the discrimination criterion can be adjusted upward or downward depending on various factors and the need to increase the sensitivity or selectivity of the test.

[00208] In some embodiments, instead of using the empirical option of deriving a recognition criterion from the number of large-volume reads from control samples, or in addition to it, a recognition criterion for determining a repeat expansion can be theoretically derived. The expected number of reads that are entirely within the range of the repeat can be calculated taking into account a number of parameters, including the length of paired-end reads, the length of the sequence having the repeat expansion, and the sequencing depth. For example, the sequencing depth can be used to calculate the average distance between reads in an aligned genome. In the case where an individual sample was sequenced at 30-fold depth, the total number of bases sequenced is equal to the product of the genome size and the depth. In the case of applying the present method to a human, this number would be about 3×10 ^{9 ×} 30=9×10 ¹⁰ . If each read is 100 bp long, This depth requires only 9×10 ⁸ reads. Since the genome is diploid, half of these reads represent sequencing of one chromosome/haplotype, and the remaining reads represent sequencing of the other chromosome/haplotype. There are 4.5×10 ⁸ reads per haplotype, and dividing the total genome size by this number yields the average distance between the starting positions of each read, i.e., on average, 3×10 ^{9 /} 4.5×10 ⁸ = 1 read for every 6.7 bp. This number can be used to estimate the number of reads that will lie entirely within the repeat sequence, based on the size of the repeat sequence in a given subject. If the total repeat sequence size is 300 bp, any read starting within the first 200 bp will be within the first 200 bp. of a given repeat sequence will be entirely within the repeat sequence (reads starting within the last 100 bp will be at least partially outside the repeat sequence based on the lengths of the 100 bp reads). Since a read is expected to align every 6.7 bp, 200 bp / (6.7 bp per read) = 30 reads are expected to align entirely within the repeat sequence. Although this number can vary, it provides an estimate of the total number of reads that will be entirely within the repeat sequence for any expansion size. The lengths of the repeat sequences and the expected numbers of reads fully aligned within the repeat sequence calculated according to this method are given in Table 2 of Example 1 below.

[00209] In some embodiments, the recognition criterion is calculated based on the distance between the first and last observations of the repeat sequence within the reads, thus allowing for mutations in the repeat sequence and sequencing errors.

[00210] In some embodiments, the method may further include diagnosing whether the subject from whom the test sample is obtained is at increased risk of developing a genetic disease such as fragile X syndrome, ALS, Huntington's disease, Friedreich's ataxia, spinocerebellar ataxia, spinobulbar muscular atrophy, myotonic dystrophy, Machado-Joseph disease, dentato-rubro-pallido-Lewis atrophy, etc. Such a diagnosis may be based on determining the likelihood of the presence of a repeat expansion in the test sample, as well as the sequence of the repeats and genes associated with the repeat expansion. In other embodiments, in cases where the genetic disease is unknown, some embodiments may be used to detect an abnormally large number of repeats to identify genetic predispositions to the disease.

[00211] Fig. 4 is a flow chart illustrating another method for detecting a repeat expansion in accordance with some embodiments. Instead of large volume reads, process 400 uses the number of repeats in paired end reads of a test sample to determine the presence of a repeat expansion. Process 400 begins by sequencing a test sample containing a nucleic acid to obtain paired end reads. See block 402, equivalent to block 302 of process 300. Process 400 continues by aligning the paired end reads to a reference sequence containing a repeat sequence. See block 404, equivalent to block 304 of process 300. The process continues by determining base values and base reads in the paired end reads, wherein the base reads are in alignment with or near the repeat sequence and the anchor reads are unaligned reads that are paired with the base reads. In some embodiments, unaligned reads include both reads that cannot be aligned and reads that are poorly aligned with the reference sequence.

[00212] After determining the base and anchored reads, process 400 obtains a repeat count in the base and/or anchored reads of the test sample. See block 408. The process then obtains a repeat count distribution in the base and/or anchored reads of the test sample. In some embodiments, only the repeat count obtained in the anchored reads is analyzed. In other embodiments, both anchored reads and base reads repeats are analyzed. The repeat count distribution of the test sample is then compared to the repeat count distribution of one or more control samples. See block 410. In some embodiments, the process determines the presence of a repeat expansion of a repeat sequence in the test sample if the distribution of the test sample is statistically different from the distribution of the control samples. See block 412. Process 400 analyzes the number of repeats relative to reads, including high-volume reads and low-volume reads, which differs from the process in which only high-volume reads are analyzed, as described above with respect to process 300.

[00213] In some embodiments, comparing the distribution of the test sample to the distribution of the control samples includes using a Mann-Whitney rank test to determine whether the two distributions differ significantly. In some embodiments, the assay determines whether a repeat expansion is likely to be present in the test sample if the distribution of the test sample is more skewed toward a higher number of repeats than the control samples and the p-value of the Mann-Whitney rank test is less than about 0.0001 or 0.00001. If necessary, the p-value may be adjusted to improve the selectivity or sensitivity of the test.

[00214] The repeat expansion detection processes described above with respect to Figs. 2-4 employ anchored reads, which are unaligned reads paired with reads aligned to the repeat sequence of interest. Variations in these processes may include searching the unaligned reads for read pairs that consist almost entirely of some type of repeat sequence to detect new, previously unidentified repeat expansions that may be medically significant. This method does not allow for quantification of the exact number of repeats, but is effective in identifying extreme repeat expansions or artifacts that should be flagged for subsequent quantification. When combined with longer reads, this method can both detect and quantify repeats up to 200 bp or more within the total length.

[00215] Fig. 5 illustrates a flow chart of a process 500 in which unaligned reads that are not associated with any repeat sequence of interest are used to detect a repeat expansion. The process 500 may use unaligned reads of the entire genome to detect a repeat expansion. The process begins with sequencing a nucleic acid sample of interest to obtain paired-end reads. See block 502. The process 500 continues by aligning the paired-end reads to a reference genome. See block 504. The process then determines unaligned reads of the entire genome. Unaligned reads include paired-end reads that cannot be aligned or are poorly aligned to the reference sequence. See block 506. The process then analyzes the number of repeats of the repeat unit in the unaligned reads to determine the likelihood of a repeat expansion in the sample of interest. Such analysis may be independent of any particular repeat sequence. Such analysis may be applied to a variety of repeat units, and the number of repeats in various repeat units of the sample of interest may be compared to the number of repeat units in several control samples. The analysis may employ the methods of comparing the sample of interest to the control samples described above. In the event that the comparison reveals that the sample of interest has an abnormally high number of repeats of the repeat unit, further analysis may be performed to determine whether the sample of interest contains a repeat expansion of the particular repeat sequence of interest. See block 510.

[00216] In some embodiments, the additional analysis includes very long sequence reads that can potentially span long repeat sequences having repeat expansions that are medically significant. The reads generated as part of this additional analysis are longer than paired end reads. In some embodiments, single-molecule sequencing or synthetic sequencing using long reads is used to obtain the long reads. In some embodiments, the association between the repeat expansion and the genetic disease is known in the art. However, in other embodiments, the association between the repeat expansion and the genetic disease may not be known in the art.

[00217] In some embodiments, analyzing the number of repeats of the repeat unit in the unaligned reads of operation 510 includes performing a high-volume analysis comparable to the analysis performed in operation 308 (see FIG. 3). The analysis includes obtaining a number of high-volume reads, wherein the high-volume reads are unaligned reads with a greater number of repeats than a threshold; and comparing the number of high-volume reads of reads in the sample under study with a recognition criterion. In some embodiments, the threshold for high-volume reads is at least about 80% of the maximum number of repeats, wherein the maximum value is calculated as the ratio of the length of paired-end reads to the length of the repeat unit. In some embodiments, the high-volume reads also include reads that are paired with unaligned reads and have more repeats than the threshold.

[00218] In some embodiments, prior to further analysis of operation 510, the process further includes (a) determining paired end reads that are paired with the unaligned reads and aligned to a repeat sequence on the reference genome or in the vicinity thereof; and (b) providing the repeat sequence as a specific repeat sequence of interest to perform operation 510. Further analysis of the repeat sequence of interest may then use any of the methods in accordance with Figs. 2-4, described above.

Samples

[00219] Samples used to determine repeat expansion may include samples taken from any cell, fluid, tissue, or organ, including nucleic acids, in which it is desired to determine repeat expansion in one or more repeat sequences of interest. In some embodiments involving fetal diagnosis, it is preferable to obtain cell-free nucleic acids, such as cell-free DNA (cfDNA), from maternal body fluid. Cell-free nucleic acids, including cell-free DNA, can be obtained by a variety of methods known in the art from biological samples including, but not limited to, plasma, serous exudate, and urine (see, e.g., Fan et al., Proc Natl Acad Sci 105:16266–16271 [2008]; Koide et al., Prenatal Diagnosis 25:604–607 [2005]; Chen et al., Nature Med. 2: 1033–1035 [1996]; Lo et al., Lancet 350: 485–487 [1997]; Botezatu et al., Clin Chem. 46: 1078–1084, 2000; and Su et al., J Mol. Diagn. 6: 101–107). [2004]).

[00220] In various embodiments, nucleic acids (e.g., DNA or RNA) present in a sample can be enriched in a specific or non-specific manner prior to use (e.g., prior to preparing a library for sequencing). In the illustrative examples depicted below, DNA is used as an example of nucleic acids. Non-specific enrichment of a DNA sample refers to whole-genome amplification of genomic DNA fragments of a sample that can be used to enhance the level of the DNA sample prior to preparing a library for cDNA sequencing. Whole-genome amplification methods are known to those skilled in the art. Examples of whole-genome amplification methods include degenerate oligonucleotide primed (DOP) PCR, primer extension PCR (PEP), and multiple displacement amplification (MDA). In some embodiments, the sample may be unenriched DNA.

[00221] A sample containing nucleic acids to which the methods described herein are applied typically includes a biological sample (a "test sample") as described above. In some embodiments, nucleic acids to be tested for repeat expansion are purified or isolated by any of a number of well-known methods.

[00222] Accordingly, in certain embodiments, the sample comprises or consists essentially of a purified or isolated polynucleotide, or may comprise samples, including a tissue sample, a biological fluid sample, a cellular sample, and the like. Suitable biological fluid samples include, but are not limited to, blood, plasma, serous exudate, sweat, lacrimal fluid, sputum, urine, otic fluid, lymph, saliva, cerebrospinal fluid, lavage, bone marrow suspension, vaginal fluid, transcervical lavage, brain fluid, ascitic fluid, respiratory, intestinal, and genitourinary secretions, amniotic fluid, milk, and leukophoresis samples. In some embodiments, the sample is a sample that is readily obtainable by non-invasive procedures, such as blood, plasma, serous exudate, sweat, lacrimal fluid, sputum, urine, otic fluid, saliva, or feces. In some embodiments, the sample is a peripheral blood sample or a plasma fraction and/or serous exudate of a peripheral blood sample. In other embodiments, the biological sample is material obtained by performing a biopsy, taking a wash, smear, or a cell culture. In another embodiment, the sample is a mixture of two or more biological samples; for example, the biological sample may include two or more samples of a biological fluid, tissue, or cell culture. As used herein, the terms "blood,""plasma," and "serous exudate" refer to fractions or processed portions thereof. Similarly, if the sample is obtained by performing a biopsy, taking a wash, smear, etc., the term "sample" refers to a processed fraction or portion obtained by performing a biopsy, taking a wash, smear, etc.

[00223] In some embodiments, samples may be obtained from sources including, but not limited to, samples obtained from different subjects, samples from the same or different subjects at different stages of development, samples from different diseased subjects (e.g., subjects suspected of having a genetic disease), healthy subjects, samples obtained at different stages of a patient's disease, samples obtained from a patient undergoing different treatments for a disease, samples obtained from patients exposed to different environmental factors, samples obtained from patients predisposed to the manifestation of pathologies, samples obtained from patients exposed to an infectious disease agent, and the like.

[00224] In one exemplary, non-limiting embodiment, the sample is a sample obtained from a pregnant woman. In this case, the sample can be analyzed using the methods described herein to perform a prenatal diagnosis of potential fetal chromosomal abnormalities. The sample obtained from the mother can be a tissue sample, a biological fluid sample, or a cellular sample. The term "biological fluid" includes, but is not limited to, blood, plasma, serous exudate, sweat, tears, sputum, urine, ear fluid, lymph, saliva, cerebrospinal fluid, lavage, bone marrow suspension, vaginal fluid, transcervical lavage, brain fluid, ascitic fluid, respiratory, intestinal and genitourinary secretions, amniotic fluid, milk and white blood cells.

[00225] In some embodiments, samples may also be obtained from tissues, cells, or other polynucleotide-containing sources cultured in vitro. Cultured samples may be obtained from sources including, but not limited to, different cultures (e.g., tissues or cells) maintained in different environments and conditions (e.g., pH, pressure, and temperature), cultures (e.g., tissues or cells) maintained for different periods of time, cultures (e.g., tissues or cells) treated with different factors or reagents (e.g., potential drugs or modulators), or cultures belonging to different types of tissues and/or cells.

[00226] Methods for isolating nucleic acids from biological sources are well understood and may vary depending on the properties of the particular source. Those skilled in the art can perform the isolation of nucleic acids from a source as needed within the framework of the method described herein. In some cases, fragmentation of nucleic acid molecules in a nucleic acid sample may be preferred. Fragmentation may be performed in a random order, as well as in a specific order, such as by restriction endonuclease digestion. Random fragmentation methods performed in a random order are well known in the art and include, for example, limited deoxyribonuclease digestion, alkaline treatment, and physical hydrodynamic fragmentation.

Preparing libraries for sequencing

[00227] In various embodiments, sequencing can be performed on a variety of sequencing platforms that require preparation of a sequencing library. Preparation typically includes DNA shearing (sonication, pulverization, or hydrodynamic shearing) followed by DNA repair and end treatment (blunt end or overhanging A end) and adaptor ligation specific to the platform used. In one embodiment, the methods described herein can utilize next generation sequencing (NGS) technologies that allow multiple samples to be sequenced individually as genomic molecules (i.e., single-channel sequencing) or as pooled samples containing indexed genomic molecules (e.g., multi-channel sequencing) in a single sequencing run. These methods can generate up to several hundred million DNA sequence reads. In various embodiments, genomic nucleic acid sequences and/or indexed genomic nucleic acids may be determined, for example, using next-generation sequencing (NGS) technologies described herein. In various embodiments, analysis of large amounts of sequence data obtained using NGS may be performed using one or more processors as described herein.

[00228] In various embodiments, the use of such sequencing technologies does not involve the preparation of libraries for sequencing.

[00229] However, in certain embodiments, the sequencing methods provided herein include preparing sequencing libraries. In one exemplary approach, preparing sequencing libraries includes obtaining a random set of adapter-modified DNA fragments (e.g., polynucleotides) ready for sequencing. Polynucleotide sequencing libraries can be prepared from DNA or RNA, including DNA or cDNA equivalents, analogs, such as DNA or cDNA that are complementary, or from replicate DNA prepared from an RNA template by the action of a reverse transcriptase. Polynucleotides can be formed in a double-stranded form (e.g., dsDNA, such as genomic DNA fragments, cDNA, PCR amplified products, etc.) or, in certain embodiments, polynucleotides can be formed in a single-stranded form (e.g., ssDNA, RNA, etc.) and converted to a dsDNA form. As an example, in certain embodiments, single-stranded mRNA molecules can be copied into double-stranded cDNAs suitable for use in preparing a library for sequencing. The exact sequence of the primary polynucleotide molecules is generally not a material for use in the library preparation method, and may or may not be known. In one embodiment, the polynucleotide molecules are DNA molecules. More specifically, in certain embodiments, the polynucleotide molecules are the entire genetic complement of an organism, or substantially the entire genetic complement of an organism, and are genomic DNA molecules (e.g., cellular DNA, cell-free DNA (cfDNA), etc.), which typically include both an intronic sequence and an exonic sequence (coding sequence), as well as non-coding regulatory sequences such as promoter and enhancer sequences. In some embodiments, the primary polynucleotide molecules comprise human genomic DNA molecules, such as cDNA molecules present in the peripheral blood of a pregnant patient.

[00230] The generation of sequencing libraries for some NGS sequencing platforms is facilitated by the use of polynucleotides containing a defined range of fragment sizes. The generation of such libraries typically involves fragmenting large polynucleotides (e.g., cellular genomic DNA) to obtain polynucleotides in the desired size range for repeat expansion determination.

[00231] The length of a fragment or insert is greater than the length of a read and is typically greater than the sum of the lengths of two reads.

[00232] In some exemplary embodiments, the nucleic acid sample(s) is/are obtained as genomic DNA that is fragmented into fragments containing about 100, 200, 300, 400, 500 or more base pairs and to which NGS methods can be applied. In some embodiments, paired-end reads are obtained from inserts that are about 100-5,000 bp in length. In some embodiments, the inserts are about 100-1,000 bp in length. Sometimes, these are implemented as regular paired-end reads with short inserts. In some embodiments, the inserts are about 1,000-5,000 bp in length. Sometimes, these are implemented as mate paired reads with long inserts, as described above.

[00233] In some embodiments, long inserts can be used to evaluate very long extended repeat sequences. In some embodiments, mate-pair reads can be used to obtain reads separated by thousands of base pairs. In these embodiments, the inserts or fragments are in the range of hundreds to thousands of base pairs with two biotin compound adapters at the two ends of the insert. The biotin compound adapters then connect the two ends of the insert to form a circular molecule, which is then further fragmented. A subfragment including the biotin compound adapters and the two ends of the original insert is selected for sequencing on a platform capable of sequencing shorter fragments.

[00234] Fragmentation can be accomplished by any of a number of methods known to those skilled in the art. For example, fragmentation can be accomplished by mechanical means, including, but not limited to, pulverization, sonication, and hydrodynamic fragmentation. However, mechanical fragmentation typically cleaves the DNA backbone at the C-O, P-O, and C-C bonds to produce a heterogeneous mixture of blunt and 3'- and 5'-overhanging ends with broken C-O, P-O, and C-C bonds (see, e.g., Alnemri and Liwack, J Biol. Chem 265:17323-17333 [1990]; Richards and Boyer, J Mol Biol 11:327-240 [1965]), which may require repair because they may lack the necessary 5'-phosphate to carry out subsequent enzymatic reactions, such as ligation of sequencing adapters needed to generate DNA for sequencing.

[00235] In contrast, scDNA typically exists in fragments of less than about 300 base pairs, and therefore fragmentation is generally not required to generate a sequencing library using scDNA samples.

[00236] In general, whether polynucleotides have been artificially fragmented (e.g., fragmented in vitro) or exist as fragments in vivo, they are converted to blunt-ended DNA containing 5'-phosphates and a 3'-hydroxyl. Standard protocols, such as sequencing protocols using, among other things, the Illumina platform as described elsewhere herein, provide instructions for end repair of the DNA sample, purification of end-repaired products prior to dA end-joining, and purification of dA end-joined products prior to library preparation steps using adapter ligation.

[00237] The various embodiments of the methods for preparing sequencing libraries described herein eliminate the need for one or more steps typically required in standard protocols to produce a modified DNA product that can be sequenced by NGS. The abbreviated method (abbreviated method), 1-step method, and 2-step method are examples of methods for preparing a sequencing library that can be found in patent application Ser. No. 13/555,037, filed July 20, 2012, the contents of which are incorporated herein by reference in their entirety.

5.12. Sequencing methods

[00238] As noted above, the resulting samples (e.g., sequencing libraries) are sequenced using a method to determine copy number variation(s). Any of a variety of sequencing technologies may be used.

[00239] Sequencing devices and technologies such as the hybridization sequencing platform manufactured by Affymetrix Inc. (Sunnyvale, CA, USA), synthesis sequencing platforms manufactured by 454 Life Sciences (Bradford, CT, USA), Illumina/Solexa (San Diego, CA, USA), and Helicos Biosciences (Cambridge, MA, USA), and the ligation sequencing platform manufactured by Applied Biosystems (Foster City, CA, USA) are commercially available, as described below. In addition to single-molecule sequencing, which is performed using the Helicos Biosciences sequencing-by-synthesis method, other single-molecule sequencing technologies include, but are not limited to, Pacific Biosciences' SMRT™ technology, ION TORRENT™ technology, and nanopore sequencing such as that developed by Oxford Nanopore Technologies.

[00240] Although the automated Sanger method is considered a "first generation" technology, Sanger sequencing, including automated Sanger sequencing, can also be used in the methods described herein. Additional sequencing methods include, but are not limited to, nucleic acid imaging technologies such as atomic force microscopy (AFM) or transmission electron microscopy (TEM). Illustrative examples of sequencing technologies are described in more detail below.

[00241] In some embodiments, the disclosed methods include obtaining nucleic acid sequence information in a test sample by massively parallel sequencing of millions of DNA fragments using Illumina sequencing by synthesis and reversible terminator-based sequencing chemistry (e.g., as described in Bentley et al., Nature 6:53-59 [2009]). The template DNA may be genomic DNA, such as cellular DNA or scDNA. In some embodiments, genomic DNA from isolated cells is used as a template and is then fragmented to form lengths of several hundred base pairs. In other embodiments, scDNA is used as a template and fragmentation is not required because the scDNA exists in short fragments. For example, fetal scDNA circulates in the bloodstream in fragments of approximately 170 base pairs (bp) in length (Fan et al., Clin Chem 56:1279–1286 [2010]), and fragmentation of the DNA is not required prior to sequencing. Illumina sequencing technology is based on affixing fragmented genomic DNA to a transparent, flat surface bearing oligonucleotide anchors. The template DNA is end-reduced to produce 5′-phosphorylated blunt ends, and the polymerase action of the Klenow fragment is used to add a single base to the 3′ end of the blunt-phosphorylated DNA fragments. This addition prepares the DNA fragments for ligation using oligonucleotide adapters that have a single T overhang at the 3′ end to improve ligation efficiency. The adapter oligonucleotides are complementary to the base oligonucleotides of the flow cell (not to be confused with the base/anchored reads in the repeat expansion assay). Under limiting and dilution conditions, adapter-modified single-stranded template DNA is added to the flow cell and immobilized by hybridization with the base oligonucleotides. Extension and bridge amplification of the added DNA fragments is performed to create an ultra-high-density sequencing flow cell with hundreds of millions of clusters, each containing about 1,000 copies of the same template. In one embodiment, randomly fragmented genomic DNA is amplified using PCR before it is subjected to cluster amplification. Alternatively, an unamplified genomic library preparation method is used and randomly fragmented genomic DNA is enriched using cluster amplification alone (Kozarewa et al., Nature Methods 6:291–295 [2009]). Arrays are sequenced using deep four-color DNA sequencing by synthesis technology that utilizes reversible terminators with removable fluorescent dyes. Highly sensitive fluorescence detection is accomplished using laser excitation and a full set of internal optical reflectors. Short reads of about tens to hundreds of base pairs are aligned to a reference genome, and the unique mapping of the short reads to the reference genome is determined using custom-designed data analysis software. After completion of the first read, arrays can be reconstructed in situ to allow a second read to be performed from the opposite end of the fragments. Thus, single or paired end sequencing of DNA fragments can be used.

[00242] In various embodiments of the present disclosure, sequencing by synthesis can be used to provide paired-end sequencing. In some embodiments, the Illumina sequencing by synthesis platform has a fragment clustering feature. Clustering is a process in which each molecule of a fragment is isothermally amplified. In some embodiments, as an example described herein, a fragment has two different adapters attached to the two ends of the fragment; such adapters allow the fragment to hybridize to two different oligonucleotides on the surface of a flow cell strip. The fragment further includes or is linked to two index sequences at the two ends of the fragment, wherein the index sequences allow the use of tags to identify different samples in multi-channel sequencing. In some sequencing platforms, a fragment to be sequenced is also referred to as an insert.

[00243] In some embodiments, a clustering flow cell used on the Illumina platform is a banded glass slide. Each band is a glass channel coated with two types of oligonucleotides. Hybridization is performed using the first of the two types of oligonucleotides located on the surface. This oligonucleotide is complementary to the first adapter at one end of the fragment. The polymerase creates a complementary strand of the hybridized fragment. The double-stranded molecule is denatured, and the original template strand is washed out. The remaining strand is clonally amplified by using bridges in parallel with the remaining strands.

[00244] In bridge amplification, a second adapter region at the second end of the strand hybridizes to a second type of oligonucleotide on the surface of the flow cell. The polymerase creates a complementary strand, forming a double-stranded bridge molecule. This double-stranded molecule is denatured, resulting in two single-stranded molecules attached to the flow cell by two different oligonucleotides. The process is then repeated a number of times; this process is carried out simultaneously with millions of clusters, resulting in clonal amplification of all fragments. After bridge amplification, the reverse strands are cleaved and washed out, leaving only the forward strands. The 3' ends are blocked to prevent unwanted priming.

[00245] After clustering, the sequencing process begins with the extension of the first sequencing primer, resulting in the first read. Within each cycle, fluorescently labeled nucleotides are recruited to the growing strand. Based on the template sequence, only one of them is inserted. After each nucleotide is added, the cluster is excited by a light source and a characteristic fluorescent signal is emitted. The number of cycles determines the length of the read. The wavelength of the emission and the intensity of the signal determine the process of base recognition. For a given cluster, reads of all identical strands are performed simultaneously. Hundreds of millions of clusters are sequenced by massively parallel sequencing. After the first read is complete, the read product is washed out.

[00246] In the next step of the work, according to the protocols using two index primers, index primer 1 is introduced into index region 1 of the template and hybridized thereto. Index regions allow for the detection of fragments that can be used to demultiplex samples as part of the multichannel sequencing process. Index read 1 is generated similarly to the first read. After the index read 1 is completed, the read product is washed out and the 3' end of the strand is deprotected. The template strand is then folded and bound to a second oligonucleotide on the flow cell. Index sequence 2 is read in the same manner as index 1. Then, upon completion of the step, the index read product 2 is washed out.

[00247] After the two reads, read 2 begins by using polymerases to expand the oligonucleotides of the second flow cell, resulting in the formation of a double-stranded bridge. This double-stranded DNA is denatured and the 3' end is blocked. The original forward strand is cleaved and washed away, leaving only the reverse strand. Read 2 begins by introducing the read 2 sequencing primer. As with read 1, the sequencing steps are repeated until the desired length is reached. Read 2 is washed away. This process can generate millions of reads representing all fragments. Sequences from the pooled sample libraries are separated based on unique indices assigned during sample preparation. For each sample, reads of similar stretches of base call results are locally clustered. Reads in the forward and reverse directions are paired, resulting in coherent sequences. These linked sequences are aligned with a reference genome to identify variants.

[00248] An example of sequencing by synthesis includes the use of paired-end reads, which are used in many embodiments of the described methods. In paired-end sequencing, 2 reads are produced from either end of a fragment. Paired-end reads are used to disambiguate alignments. When performing paired-end sequencing, users can select the length of an insert (or fragment to be sequenced) and sequence both ends of the insert, resulting in the generation of high-quality aligned sequence data. Since the distance between all paired-end reads is known, alignment algorithms can use this information to more accurately align repetitive regions. This allows for better alignment of reads, especially in difficult-to-sequence repetitive regions of the genome. Using paired-end sequencing, rearrangements can be detected, including insertions, deletions, and inversions.

[00249] For paired-end reads, inserts of different lengths (i.e., different fragment sizes for sequencing) can be used. As a default value, paired-end reads are used herein to refer to reads obtained from different insert lengths. In some cases, paired-end reads with short inserts are distinguished from paired-end reads with long inserts, the latter in particular are referred to as mate-pair reads. In some embodiments comprising mate-pair reads, two biotin compound adapters are first attached to the two ends of a relatively long insert (e.g., several kb in length). The biotin compound adapters then link the two ends of the insert to form a circular molecule. A subfragment combining the biotin compound adapters can be obtained by further fragmenting the circular molecule. The subfragment, which includes the two ends of the original fragment in opposite sequence order, can then be sequenced using the same procedure as for the short insert paired end sequencing described above. Additional information on performing mate pair sequencing using the Illumina platform is provided in the online publication at the following address: res.illumina.com/documents/products/technotes/technote_nextera_matepair_data_processing.pdf. This publication is also incorporated herein by reference in its entirety.

[00250] After sequencing DNA fragments, sequence reads of a given length, such as 100 bp, are aligned or mapped to a known reference genome. The aligned or mapped reads and their corresponding locations on the reference sequence are also referred to as tags. In many embodiments of the assays described herein, reads that are either poorly aligned or cannot be aligned at all are used to determine repeat expansion, as well as aligned reads (tags). In one embodiment, the reference genomic sequence is the NCBI36/hg18 sequence, which can be found on the Internet at the following address: genome.ucsc.edu/cgi-bin/hgGateway?org=Human&db=hg18&hgsid=166260105). In an alternative embodiment, the reference genomic sequence is the GRCh37/hg19 sequence, which can be found on the Internet at the following address: genome.ucsc.edu/cgi-bin/hgGateway. Other sources of publicly available sequence information include the Genetic Database, dbEST, dbSTS, EMBL (European Molecular Biology Laboratory), and DDBJ (DNA Data Bank of Japan). Sequence alignment can be performed using a number of computer algorithms, including, but not limited to, BLAST (Altschul et al., 1990), BLITZ (MPsrch) (Sturrock & Collins, 1993), FASTA (Person & Lipman, 1988), BOWTIE (Langmead et al., Genome Biology 10:R25.1-R25.10 [2009]), or ELAND (Illumina, Inc., San Diego, CA, USA). In one embodiment, one end of the clonally expanded copies of plasma cDNA molecules is sequenced and processed by performing bioinformatic analysis of the alignments on the Illumina Genome Analyzer using Efficient Large-Scale Alignment of Nucleotide Databases (ELAND) software.

[00251] In one exemplary, non-limiting embodiment, the methods described herein include obtaining nucleic acid sequence information in a test sample using Helicos True Single Molecule Sequencing (tSMS) technology (e.g., described in Harris, TD et al., Science 320:106-109 [2008]). In the tSMS technique, a DNA sample is sheared into strands of about 100 to 200 nucleotides in length and a polyA sequence is added to the 3' end of each DNA strand. Each strand is labeled by the addition of fluorescently labeled adenosine. The DNA strands are then hybridized to a flow cell containing millions of T-nucleotide capture sites immobilized on the surface of the flow cell. In certain embodiments, the arrays may have a density of about 100 million arrays/ ^cm2 . The flow cell is loaded into an instrument such as a HeliScopeTM sequencer and a laser is shined onto the surface of the flow cell, revealing the position of each template. A CCD camera can be used to map the positions of the templates onto the surface of the flow cell. The fluorescent template label is then cleaved and washed away. The sequencing reaction begins with the introduction of DNA polymerase and a fluorescently labeled nucleotide. The T-oligonucleotide acid serves as a primer. The polymerase incorporates the labeled nucleotides into the primer under the guidance of the template. The polymerase and unincorporated nucleotides are removed. Templates into which a fluorescently labeled nucleotide is specifically incorporated are identified by visualizing the surface of the flow cell. After imaging, the fluorescent label is removed in a digestion step, and the process is repeated with additional fluorescently labeled nucleotides until the desired read length is reached. Sequence information is collected at each nucleotide addition step. Whole-genome sequencing using single-molecule sequencing technologies eliminates the need for PCR amplification to prepare sequencing libraries, and the methods also allow direct measurement of a sample rather than measuring copies of that sample.

[00252] In another illustrative, non-limiting embodiment, the methods described herein involve obtaining nucleic acid sequence information in a test sample using 454 sequencing (Roche) (e.g., as described in Margulies, M. et al. Nature 437:376-380 [2005]). 454 sequencing typically involves two steps. In the first step, DNA is hydrodynamically sheared to separate it into fragments of about 300-800 base pairs each and blunt-ended. Oligonucleotide adapters are then ligated to the ends of the fragments. The adapters serve as primers to perform amplification and sequencing of the fragments. The fragments can be attached to DNA capture beads, such as streptavidin-coated beads, using, for example, adapter B, which contains a 5'-biotin label. The fragments attached to the beads are amplified by PCR in droplets of an oil-in-water emulsion. This results in multiple copies of the clonally amplified DNA fragments on each bead. In a second step, the beads are captured into wells (e.g., picoliter wells). Pyrosequencing of each DNA fragment is performed in parallel. The addition of one or more nucleotides generates a light signal that is recorded by a CCD camera in the sequencing apparatus. The signal strength is proportional to the number of incorporated nucleotides. Pyrosequencing uses pyrophosphate (PPi), which is released upon addition of nucleotides. PPi is converted to ATP by ATP sulfurylase in the presence of adenosine 5'-phosphosulfate. Luciferase interacts with ATP to convert luciferin to oxyluciferin, a reaction that generates light that is then measured and analyzed.

[00253] In another exemplary, non-limiting embodiment, the methods described herein include obtaining nucleic acid sequence information in a test sample using SOLiD™ technology (Applied Biosystems). In ligation sequencing using SOLiD™ technology, genomic DNA is hydrodynamically sheared to separate it into fragments, and adapters are attached to the 5' and 3' ends of the fragments to generate a library of fragments. In an alternative embodiment, internal adapters can be introduced by ligating adapters to the 5' and 3' ends of the fragments, rounding the fragments, cleaving the circular fragment to generate an internal adapter, and attaching adapters to the 5' and 3' ends of the resulting fragments to generate a mate pair library. Clonal bead populations are then prepared in microreactors containing the beads, primers, template, and PCR components. Following PCR, the templates are denatured and the beads are enriched to separate beads with extended templates. The templates on selected beads are 3' modified to establish bonds with the slide. Sequencing is accomplished by sequential hybridization and ligation of partially random oligonucleotides to a central specific base (or base pair) determined by a specific fluorophore. After color detection, the ligated oligonucleotide is cleaved and removed, and the process is repeated.

[00254] In another exemplary, non-limiting embodiment, the methods described herein include obtaining nucleic acid sequence information in a sample of interest using Pacific Biosciences' sequencing real-time (SMRT™) technology. SMRT sequencing visualizes the continuous incorporation of dye-labeled nucleotides during DNA synthesis. Single DNA polymerase molecules are attached to the bottom surface of individual zero-mode wavelength detectors (ZMW detectors) that obtain sequence information during the incorporation of phospho-linked nucleotides into a growing primer strand. The ZMW detector contains a limiting structure that allows the incorporation of a single nucleotide by DNA polymerase to be observed against a background of fluorescent nucleotides that rapidly diffuse into and out of the ZMW (within microseconds). Incorporation of a nucleotide into a growing strand typically takes several milliseconds. During this time, the fluorescent label is excited and produces a fluorescent signal, after which the fluorescent label is cleaved. Measuring the corresponding fluorescence of the dye indicates which base has been inserted. The process is repeated to obtain the sequence.

[00255] In another illustrative, non-limiting embodiment, the methods described herein include obtaining nucleic acid sequence information in a sample of interest using nanopore sequencing (e.g., as described in Soni GV and Meller A. Clin Chem 53: 1996-2001 [2007]). Nanopore sequencing techniques for DNA analysis have been developed by a number of companies, including, for example, Oxford Nanopore Technologies (Oxford, UK), Sequenom, NABsys, and the like. Nanopore sequencing is a single-molecule sequencing technology in which a single DNA molecule is sequenced directly as it passes through a nanopore. A nanopore is a small hole, typically on the order of 1 nanometer in diameter. Immersion of a nanopore in a conductive fluid and connection of a potential (voltage) source to it results in the appearance of a small electric current due to the conduction of ions through the nanopores. The strength of the current depends on the size and shape of the nanopore. As a DNA molecule passes through a nanopore, each nucleotide of the DNA molecule blocks the nanopore to a different extent, changing the magnitude of the current passing through the nanopore to a different extent. Thus, such a change in current as a DNA molecule passes through a nanopore allows for the reading of the DNA sequence.

[00256] In another exemplary, non-limiting embodiment, the methods described herein include obtaining nucleic acid sequence information in a sample under test using a chemically sensitive field-controlled transistor (chemFET) (e.g., as described in U.S. Patent Application Publication No. 2009/0026082). In one embodiment of this technique, DNA molecules can be placed in reaction chambers and template molecules can be hybridized to a sequencing primer linked to a polymerase. The incorporation of one or more triphosphates into a new nucleotide chain at the 3' end of the sequencing primer can be detected by the chemFET as a change in current. The array can include a plurality of chemFET sensors. In another example, individual nucleic acids can be attached to microspheres, nucleic acid amplification can be performed on the microsphere, and the individual microspheres can be transferred to individual reaction chambers on a chemFET array, each of the chambers being equipped with a chemFET sensor; in addition, nucleic acid sequencing can be performed.

[00257] In another embodiment, the DNA sequencing technology is single-molecule sequencing using Ion Torrent technology, which combines semiconductor technology with the chemistry of conventional sequencing to directly convert chemically encoded information (A, C, G, T) into digital information (0, 1) using a semiconductor chip. In nature, when a nucleotide is inserted into a DNA strand by a polymerase, a hydrogen ion byproduct is released. Ion Torrent technology uses a high-density array of micromachined wells to perform this biochemical process by massively parallel sequencing. Each well contains an individual DNA molecule. Under the wells is an ion-sensitive layer, and under it is an ion sensor. When a nucleotide, such as C, is added to the DNA template and then inserted into the DNA strand, a hydrogen ion is released. The charge of this ion will change the pH of the solution, which in turn is recorded by the Ion Torrent ion sensor. The sequencer, which is the world's smallest solid-state pH meter, is used to recognize the base, converting chemical analysis data into digital data. The Ion Personal Genome Machine (PGM™) sequencer then sequentially fills the chip with nucleotides one after another. If the next nucleotide filling the chip does not match, the voltage changes will not be recorded, and the bases will not be recognized. If there are two identical bases on the DNA strand, the voltage doubles, and the chip will register two identical bases. The direct detection method allows for the registration of nucleotide incorporation within a few seconds.

[00258] In another embodiment, the present method includes obtaining nucleic acid sequence information in a test sample using sequencing by hybridization. Sequencing by hybridization comprises contacting a plurality of polynucleotide sequences with a plurality of polynucleotide probes, wherein each of the plurality of polynucleotide probes may optionally be attached to a substrate. The substrate may be a flat surface containing an array of known nucleotide sequences. A pattern may be used to perform hybridization with the array to determine the polynucleotide sequences present in the sample. In other embodiments, each probe is attached to a bead, such as a magnetic bead, etc. Hybridization of the bead may be used to determine the plurality of polynucleotide sequences in the sample.

[00259] In some embodiments of the methods described herein, the sequence reads are about 20 bp, about 25 bp, about 30 bp, about 35 bp, about 40 bp, about 45 bp, about 50 bp, about 55 bp, about 60 bp, about 65 bp, about 70 bp, about 75 bp, about 80 bp, about 85 bp, about 90 bp, about 95 bp, about 100 bp, about 110 bp, about 120 bp, o., about 130, about 140 bp, about 150 bp, about 200 bp, about 250 bp, about 300 bp, about 350 bp, about 400 bp, about 450 bp, or about 500 bp. It is expected that technological advances in the generation of paired-end reads will enable single end reads greater than 500 bp and reads greater than 1,000 bp. In some embodiments, paired-end reads that contain sequence reads of lengths from about 20 bp to 1,000 bp, from about 50 bp to 1,000 bp, are used to detect repeat expansions. up to 500 bp or from 80 bp to 150 bp. In various embodiments, paired end reads are used to evaluate a sequence having a repeat expansion. The sequence having a repeat expansion is longer than the reads. In some embodiments, the sequence having a repeat expansion is longer than about 100 bp, 500 bp, 1,000 bp, or 4,000 bp. Alignment of the sequence reads is accomplished by comparing the sequence of the reads to the sequence of a reference sample to determine the chromosomal origin of the sequenced nucleic acid molecule, and in such a case, specific genetic sequence information is not required. A small degree of mismatch (0-2 mismatches per read) is allowed to account for minor polymorphisms that may exist between the reference genome and the genomes in the mixed sample. In some embodiments, reads that are aligned to the reference sequence are used as base reads, and those reads that are paired with base reads but cannot be aligned or are poorly aligned to the reference sequence are used as anchor reads. In some embodiments, poorly aligned reads may have a relatively large number of mismatches per read, such as at least about 5%, 10%, 15%, or 20% mismatches per read.

[00260] Typically, a plurality of sequence tags (i.e., reads aligned to a reference sequence) are obtained for each sample. In some embodiments, at least about 3×10 ⁶ sequence tags, at least about 5×10 ⁶ sequence tags, at least about 8×10 ⁶ sequence tags, at least about 10×10 ⁶ sequence tags, at least about 15×10 ⁶ sequence tags, at least about 20×10 ⁶ sequence tags, at least about 30×10 ⁶ sequence tags, at least about 40×10 ⁶ sequence tags, or at least about 50×10 ⁶ sequence tags, comprising, for example, 100 bp, are obtained by aligning reads to a reference genome relative to each sample. In some embodiments, all sequence reads are mapped to all regions of a reference genome, allowing reads to be performed across the entire genome. In other embodiments, reads are mapped to a sequence of interest, such as a chromosome, a segment of a chromosome, or a repeat sequence of interest.

5.13. Device and systems for determining repeat expansion

[00261] The analysis of sequencing data and diagnostics obtained therefrom are typically performed using various computer algorithms and programs. Thus, in some embodiments, processes are used that include the use of data stored or transmitted by one or more computer systems or other processing systems. The embodiments described herein also relate to a device used to perform these operations. This device may be specially designed to achieve the desired goals, but such a device may also be a general-purpose computer (or a group of computers) specially configured by a computer program and / or a data structure stored in the computer. In some embodiments, a group of processors is responsible for performing some or all of these analytical operations in a joint (e.g., via network or cloud computing) and / or parallel manner. The processor or group of processors used to implement the methods described herein may be of various types, including microcontrollers and microprocessors, such as programmable devices (e.g., FPGAs and FPGAs), and non-programmable devices, such as specialized gate array ICs or general-purpose microprocessors.

[00262] In one embodiment, a system is provided for determining genotypes of variants at genomic loci, including repeat sequences, the system comprising a sequencer for receiving nucleic acid samples and providing nucleotide sequence information of such samples; a processor; and a machine-readable storage medium containing instructions executable on the processor and used to genotype variants using the following methods: (a) collecting nucleic acid sequence reads of a test sample from a database; (b) aligning the sequence reads with one or more repeat sequences, each of which is represented in a sequence graph, wherein the sequence graph has a directed graph data structure, wherein the nodes represent nucleotide sequences and directed edges connect the nodes, and wherein the sequence graph comprises one or more proper simple cycles, wherein each proper simple cycle represents a subsequence of repeats, wherein each subsequence of repeats comprises repeats of a repeating unit of one or more nucleotides; and (c) determining one or more genotypes of one or more repeat sequences using sequence reads aligned to the one or more repeat sequences.

[00263] In some embodiments of any of the systems provided herein, the sequencer is configured to perform next generation sequencing (NGS). In some embodiments, the sequencer is configured to perform massively parallel sequencing using sequencing by synthesis using reverse dye terminators. In other embodiments, the sequencer is configured to perform sequencing by ligation. In other embodiments, the sequencer is configured to perform single-molecule sequencing.

[00264] Additionally, certain embodiments relate to tangible media and/or computer-readable media intended for long-term storage of information, or computer program products that include program instructions and/or data (including data structures) for performing various operations using a computer. Examples of computer-readable media include, but are not limited to, semiconductor memory devices, magnetic media such as disk drives, magnetic tape, optical media such as disks, magneto-optical media, and hardware devices specially provided for storing and executing program instructions, such as read-only memories (ROM) and random access memories (RAM). Computer-readable media may be directly or indirectly controlled by an end user. Examples of directly controlled media include media located at the user's facility and/or media from which data is not publicly available. Examples of indirectly controlled media include media indirectly accessible to a user through an external network and/or a public service, such as a "cloud service." Examples of software instructions include both machine code, including that produced by a compiler, and files containing higher-level code that can be executed by a computer using an interpreter.

[00265] In various embodiments, the data used in the described methods, as well as used in the device, are in an electronic format. Such data may include reads and tags obtained from a nucleic acid sample, reference sequences (including reference sequences that exclusively or predominantly result in the formation of polymorphisms), recognitions such as repeat expansion recognitions, recommendations, contraindications, diagnoses, etc. In accordance with the content of this document, the data in electronic format can be stored in the device and transmitted between devices. Typically, the data in electronic format is provided in digital form and can be stored as a set of bits and/or bytes in various data structures, lists, databases, etc. The data can be implemented in electronic, optical, or other forms.

[00266] In one embodiment, a computer program product is provided for generating output data indicating the presence or absence of a repeat expansion in a test sample. The computer product may comprise instructions for implementing any one or more of the above-described methods for determining a repeat expansion. As explained above, the computer product may include a tangible computer-readable medium and/or a medium intended for long-term storage of information, containing recorded executable or compiled program code (e.g., instructions) that enables a processor to perform the determination of pinned reads and repeats in the pinned reads, as well as the presence or absence of a repeat expansion. In one example, the computer product comprises a computer-readable medium containing recorded executable or compiled program code (e.g., instructions) that enables a processor to perform the determination of a repeat expansion, and comprising: a procedure for obtaining sequencing data from at least a portion of nucleic acid molecules aligned with a repeat sequence; a machine algorithm for performing repeat expansion analysis based on said received data; and a procedure for generating output data indicating the presence, absence and type of said repeat expansion.

[00267] Sequence information in a sample of interest may be compared to reference chromosomal sequences to determine paired end reads aligned with or anchored to a repeat sequence of interest and to determine the repeat expansion of the repeat sequence. In various embodiments, the reference sequences are stored in a database, such as a relational or object-oriented database.

[00268] It should be understood that in most cases, it is impractical or even impossible to perform the computational operations within the methods described herein without assistance. For example, mapping a single read from a 30 bp sample to any of the human chromosomes can take several years if a computing device is not used. Of course, the problem is compounded by the fact that good repeat expansion recognition requires mapping thousands (e.g., at least about 10,000) or even millions of reads to a single chromosome or more.

[00269] In various embodiments, raw sequence reads are aligned to one or more sequence graphs representing one or more sequences of interest. In various embodiments, at least 10,000, 100,000, 500,000, 1,000,000, 5,000,000, or 10,000,000 reads are aligned to the one or more sequence graphs. In various embodiments, the one or more sequence graphs include at least 1, 2, 5, 10, 50, 100, 500, 1,000, 5,000, 10,000, or 50,000 sequence graphs.

[00270] In some embodiments, raw sequence reads are initially aligned to a reference genome to determine genomic coordinates of the reads before a subset of the initially aligned reads are aligned to one or more sequence graphs representing one or more sequences of interest. In various embodiments, at least 10,000, 100,000, 500,000, 1,000,000, 5,000,000, 10,000,000, or 100,000,000 reads are initially aligned to the reference genome. In some embodiments, the initially aligned reads are re-aligned to the sequence graphs to determine repeat expansions across multiple regions (each region corresponding to a sequence graph). The total number of reads that are re-aligned to the sequence graphs during each application of embodiments may range from thousands to many millions of reads. In various embodiments, 10,000,000 reads are re-aligned to each sequence graph. In various embodiments, the one or more sequence graphs include at least 1, 2, 5, 10, 50, 100, 500, 1,000, 5,000, 10,000, or 50,000 sequence graphs.

[00271] The methods described herein can be implemented using a system for determining genotypes of variants at a genomic locus that includes a repeat sequence. The system can include: (a) a sequencer for receiving nucleic acids from a test sample that provides information about the nucleotide sequence of the sample; (b) a processor; and (c) one or more computer-readable storage media containing instructions executable on the processor and used to genotype variants at genomic loci, including repeat sequences. In some embodiments, instructions for implementing the methods can be obtained from a computer-readable medium that stores computer-readable instructions for performing the determination of repeat expansions. Thus, in one embodiment, a computer program product is proposed, comprising a machine-readable medium intended for long-term storage of information, on which a program code is stored, which, when executed by one or more processors of a computer system, results in the implementation of a method for determining and expanding repeats of a repeat sequence in a test sample comprising nucleic acids, wherein the repeat sequence comprises repeats of a repeating unit of nucleotides. The program code may include: (a) code for collecting sequence reads of the test sample from a database; (b) code for aligning the sequence reads with one or more repeat sequences, each of which is represented in a sequence graph, wherein the sequence graph has a data structure of a directed graph, wherein the nodes represent nucleotide sequences and directed edges connecting the nodes, and wherein the sequence graph contains one or more proper simple cycles, wherein each proper simple cycle represents a subsequence of repeats, wherein each subsequence of repeats comprises repeats of a repeating unit of one or more nucleotides; and (c) code for determining one or more genotypes of one or more repeat sequences using sequence reads aligned to the one or more repeat sequences.

[00272] In some embodiments, the instructions may further include automatically recording information related to the method, such as repeats and fixed reads, as well as the presence or absence of indications of the presence of repeat expansions in the medical records of the human subject from whom the test sample is collected. The patient's medical records may be maintained and stored, for example, in a laboratory, a physician's office, a hospital, a healthcare facility, an insurance company, or the method of using a personal medical record may further include prescribing, initiating, and/or changing treatment for the patient from whom the test sample is collected. This process may include performing one or more additional tests or analyses on additional samples collected from the subject.

[00273] The described methods can also be performed using a computer processing system specially adapted to perform the determination of repeat expansions. In one embodiment, an implementation of a computer processing system specially configured to perform a sequence of actions within the method described herein is provided. In one embodiment, the apparatus includes a sequencing device specially configured to sequence at least a portion of the nucleic acid molecules in a sample to obtain sequence type information described elsewhere herein. The apparatus can also include components for processing the sample. Such components are described elsewhere herein.

[00274] The sequence or other data may be input into the computer or stored on a computer-readable medium in a direct or indirect manner. In one embodiment, the computer system is directly connected to a sequencing device that reads and/or analyzes nucleic acid sequences obtained from samples. Obtaining sequences or other data using such devices is performed by interacting with an interface of the computer system. In an alternative embodiment, the sequences processed by the system are obtained from a sequence storage location, such as a database or other storage. After accessing the processing device, a storage device or mass storage device is used to buffer or store (including temporarily) the nucleic acid sequences. In addition, the storage device may store the number of tags for different chromosomes or genomes, etc. The storage device may also store various routines and/or programs for analyzing the sequence representation or associated data. Such programs/routines may include programs for performing statistical analysis, etc.

[00275] In one example, a user places a sample in a sequencing machine. Data collection and/or analysis is performed using a sequencing machine connected to a computer. Software installed on the computer allows data collection and/or analysis. Data may be stored, displayed (using a monitor or other similar device), and/or sent to other locations. The computer may be connected to an Internet network used to transmit data to a mobile device used by a remote user (e.g., a physician, scientist, or laboratory technician). It should be understood that the data may be stored and/or analyzed before such transmission. In some embodiments, raw data is collected and remotely sent to the user or to a data analysis and/or storage device. Data transmission may be accomplished by storing the data on a machine-readable medium, wherein such medium may be sent to the end user (e.g., by mail). The remote user may be located in the same or a different geographic location, including, without limitation, a building, city, state, country, or continent.

[00276] In some embodiments, the methods also collect data from a plurality of polynucleotide sequences (e.g., reads, tags, and/or reference chromosomal sequences) and send the data to a computer or other computing system. For example, the computer can be connected to laboratory equipment, such as a sample collection apparatus, a nucleotide amplification apparatus, a nucleotide sequencing apparatus, or a hybridization apparatus. The computer can then collect the corresponding data collected using the laboratory equipment. The data can be stored on the computer at any stage, such as during their collection in real time, before they are sent, during the sending, or after. The data can be stored on a computer-readable medium that can also be removed from the computer. The collected or stored data can be transmitted from the computer to a remote location, such as via a local area network or a wide area network such as the Internet. At the remote location, various operations can be performed on the transmitted data, as described below.

[00277] The following describes the types of electronically formatted data that can be stored, transmitted, analyzed, and/or processed in the systems, device, and methods described herein.

Reads obtained by sequencing nucleic acids in a test sample

Tags obtained by aligning reads to a reference genome or other reference sequence or sequences

Reference genome or sequence

Locus specification, which specifies the locus type, its location and structure Read coverage

Genotype of variants Sequence graph Paths of graph

Graph alignment information

Actual recognition of repeat expansion

Diagnoses (clinical condition associated with corresponding recognitions)

Recommendations for additional testing based on recognitions and/or diagnostics

Treatment and/or monitoring plans derived from recognition and/or diagnosis

[00278] These types of data may be acquired, stored, analyzed, and/or processed at one or more locations using specialized devices. There are many ways in which they may be processed. In one case, all or most of this information is stored and used at the location where the sample being tested is processed, such as in a physician's office or other clinical setting. In another case, the sample is acquired at one location, processed and optionally sequenced at another location, the reads are aligned and recognized at one or more other locations, and diagnoses, recommendations, and/or plans are prepared at yet another location (which may be the location where the sample was collected).

[00279] In various embodiments, reads are generated by a sequencing machine and then transmitted to a remote site where they are processed to recognize the presence of a repeat expansion. For example, at this remote location, the reads are aligned to a reference sequence to create base and anchor reads. Processing operations that may be applied at various locations include the following:

Sample collection

Sample processing prior to sequencing Sequencing

Sequence data analysis and recognition of the presence of repeat expansions

Diagnostics

Communicating the diagnosis and/or detection result to the patient or health care professional Developing a plan for further treatment, testing and/or monitoring Implementing the plan

Consulting

[00280] These operations may be automated in accordance with the content of other sections of this document. Typically, sequencing, analysis of sequence data, and recognition of the presence of repeat expansion are performed by computation. Other operations may be performed manually or automatically.

[00281] Fig. 6 illustrates one implementation of a dispersed system for performing recognition or diagnosis based on a test sample. A sample collection site 01 is used to obtain a test sample from a patient. The samples are then transferred to a processing and sequencing site 03, where the test samples are processed and sequenced as described above. At site 03 is a device for processing the sample, as well as a device for sequencing the processed sample. The result of the sequencing, as described elsewhere herein, is a set of reads that are typically in an electronic format and transmitted to a network such as the Internet; such a set is designated by the reference numeral 05 in Fig. 6.

[00282] The sequence data is transmitted to a remote location 07 where the analysis and recognition are performed. This location may contain one or more powerful computing devices, such as computers or processors. After the analysis is completed and the recognition is performed based on the received sequence information at location 07, the results of this recognition are returned to the network 05. In some embodiments, in addition to the recognition results, diagnostic information is also generated at location 07. The recognition and/or diagnostic results are then transmitted over the network and returned to the sample collection location 01, as shown in Fig. 6. As explained above, this is only one of many embodiments of how the various operations associated with the creation of recognition or diagnostic results can be divided between different locations. One common embodiment involves performing the collection, processing, and sequencing of the samples at one location. Another embodiment includes performing the processing and sequencing at the same location where the analysis was performed and the recognition results were created.

EXPERIMENTAL

Examples

[00283] The program was applied to a simulated dataset containing a wide range of CAG and CCG repeat sizes in the HTT locus. HTT or huntingtin is a disease gene associated with Huntington's disease (HD), a neurodegenerative disorder characterized by the loss of neurons in the striatum. It is thought to be caused by an enlarged unstable trinucleotide repeat in the gene called huntingtin, which is translated as a polyglutamine repeat in the protein produced. A fairly wide range of trinucleotide repeats (9-35) were detected in normal control samples, and repeat numbers greater than 40 were termed pathological.

[00284] As expected, the accuracy achieved using the methods of some embodiments was significantly higher in cases where reads were aligned to a sequence graph containing both repeats, compared to cases where both repeats were analyzed independently. A similar improvement was observed compared to another CTP genotyping process (Figure 7).

[00285] To demonstrate the ability of the methods to degenerate DNA repeat genotypes, we analyzed the polyalanine repeat in the PHOX2B gene in 150 healthy control samples and one sample containing a known pathogenic expansion. PHOX2B contains a 20-codon polyalanine repeat that can expand to cause the manifestation of congenital central hypoventilation syndrome. Consistent with the known description of the repeat (Amiel et al. 2003), all but a few control samples were genotyped at a 20/20 ratio. In the embodiments, a single sample with a 20/27 expansion was accurately genotyped; this genotype was confirmed to be correct by Sanger sequencing. Also, within the scope of embodiments, the pathogenic ONV adjacent to the adenosine homopolymer in the MSH2 gene was correctly identified in three replicates obtained from whole genome sequencing of a sample obtained from SeraCare Life Sciences (see the Supplementary Materials section).

[00286] Thus, we have developed a new method that addresses the need for more accurate genotyping of complex loci. This method enables genotyping of polyalanine repeats and separation of complex regions containing repeats in close proximity to small variants and other repeats. The flexibility of the sequence graph structure used herein is expected to enable many new variant recognition methods.

Analysis of CAG and CCG CTPs in the HTT locus

[00287] For each combination of CAG and CCG repeat genotypes, a short read pattern was simulated using WGSIM (Li, n/a). We set the read length value to 150, the mate end distance value to 350, the mate end distance standard deviation value to 50, the mutation rate and base error rate to 0.0010, and the deletion rate to 0. The number of pairs was set to obtain 40-fold coverage of the locus. Reads were aligned to the GRCh37 reference with BWA-MEM 0.7.17-r1194-dirty (Li 2013).

[00288] We analyzed these data in two ways using several implementation options. First, we specified the structure of the HTT locus using the expression (CAG)*CAACAG(CCG)*, aligning the reads to a sequence graph containing both repeats. We then used several implementation options to perform independent analyses of each repeat. In this mode, reads were aligned to a graph representing the CAG CTP, and they were also individually aligned to a graph representing the CCG CTP. Finally, we analyzed both repeats independently using the latest versions of GangSTR and TredParse, which do not use sequence graphs to align reads to the repeat region (Figure 7).

[00289] Figure 7 illustrates the genotyping accuracy of the CAG and CCG CTPs at the HTT locus based on simulated data. (a) Characteristics of an embodiment in which reads are aligned to a sequence graph containing both repeats; (b) characteristics of an embodiment in which repeats are analyzed independently of each other; (c) GangSTR characteristics; (d) TredParse characteristics. We measured the maximum percentage deviation of predicted CTP lengths from the expected CTP lengths of each simulated sample. The dashed blue lines indicate the genotype in the reference genome, and the solid red lines indicate the pathogenic expansion threshold. This embodiment accurately predicts the correct genotypes for all samples when both samples are genotyped together. Joint genotyping has a major advantage over separate genotyping of both repeats using any of the three tools, which is characterized by a high relative error rate, especially in cases where one repeat is significantly longer than the other.

Datasets

[00290] PCR-free whole-genome sequencing data from 150 unrelated control samples used for analysis of the PHOX2B polyalanine repeat constitute the Polaris Diversity cohort. This cohort consists of samples selected from the International Genome Sample Resource (1000 Genomes Project Consortium et al. 2015) (www|.|internationalgenome|.|org/). Whole-genome sequencing data are available from the European Genome-Phenome Archive (EGA; www|.|ebi|.|ac|.|uk/ega/home; PRJEB20654) and the National Center for Biotechnology Information Sequence Read Archive (SRA; www|.|ncbi|.|nlm|.|nih|.|gov/sra; bioproject:387148). Sample descriptions (github|.|com/Illumina/Polaris/wiki/HiSeqX-Diversity-Cohort).

[00291] A sample containing a 20/27 PHOX2B expansion was provided by Genetics Laboratories Molecular Genetics, Addenbrooke's Medical Research Center, Connecticut, and the University of Cambridge.

[00292] The Seraseq Inherited Cancer DNA Mix v1 sample provided by SeraCare Life Sciences contains distinct pathogenic variants that are difficult to detect. These variants were synthetically added to the well-characterized GM24385 cell line; the expected frequencies of such variants are 50%. One of these artificial mutations corresponds to the ONB in the MSH2 gene, which is immediately adjacent to the long homopolymer A region.

[00293] Three replicates of the SeraCare sample were prepared using the Illumina TruSeq PCR Free kit. The replicates were run on a single lane of NovaSeq6000 using the XP workflow and sequenced at 2 × 151 read length. The replicates were analyzed using Sentieon DNASeq FASTQ to VCF and Whole Genome Resequencing v8.0.0 on the Basespace cloud platform (https://basespace.illumina.com). Sentieon can be considered as a proxy for the variant calling performance of Broad's BWA-GATK software suite, which implements the same algorithms. None of the software units correctly called the MSH2 ONV. The Fastq files and analysis results are available in the following section of the Basespace cloud platform: https://basespace.illumina.com/s/HAQNxJyEtJLP

[00294] The contents of this description may be implemented in other forms, provided that its essential characteristics are preserved. The described embodiments should be considered only as illustrative embodiments and not limiting. Accordingly, the scope of application of this document is determined by the corresponding appended claims, and not by the above description. All changes that come within the meaning and range of equivalence of the claims should be included within their scope of application.

--->

SEQUENCE LIST

<110> Illumina, Inc.

Dolzhenko Egor

<120> A SEQUENCE GRAPH BASED TOOL FOR DETERMINING

VARIATIONS IN THE SHORT TANDEM REPEAT AREAS

<130>ILMNP037

<150> US 62/815,322

<151> 2019-03-07

<160> 5

<170> PatentIn version 3.5

<210> 1

<211> 12

<212> DNA

<213> Artificial sequence

<220>

<223> Synthesized

<400> 1

cagcaacagc cg 12

<210> 2

<211> 12

<212> DNA

<213> Artificial sequence

<220>

<223> Synthesized

<400> 2

cagcaacagc gg 12

<210> 3

<211> 14

<212> DNA

<213> Artificial sequence

<220>

<223> Synthesized

<400> 3

tcgtcggcag cgtc 14

<210> 4

<211> 15

<212> DNA

<213> Artificial sequence

<220>

<223> Synthesized

<400> 4

cagcagcagc agcag 15

<210> 5

<211> 15

<212> DNA

<213> Artificial sequence

<220>

<223> Synthesized

<400> 5

ccgccgccgc cgccg 15

<---

Claims (16)

1. A method for sequencing genomic loci comprising two or more repeat sequences, wherein the method is implemented using a computer equipped with one or more processors and system memory, for genotyping one or more repeat sequences, each of which comprises one or more repeat subsequences, comprising: (a) obtaining a sequence graph, wherein the sequence graph has a graph data structure with vertices representing nucleotide sequences and directed edges connecting the vertices, and wherein the sequence graph contains two or more proper simple cycles, wherein each proper simple cycle represents a subsequence of repeats; (b) aligning, using one or more processors, the sequence reads of the sample under study against a reference genome to determine the genomic coordinates of the sequence reads and selecting a subset of the sequence reads; and (c) aligning, using one or more processors, a selected subset of sequence reads with two or more repeat sequences represented by a sequence graph representing the genomic locus. 2. The method according to claim 1, wherein each subsequence of repeats contains repeats of a repeating unit of one or more nucleotides. 3. The method according to any one of claims 1 or 2, wherein the repeat sequence of two or more repeat sequences comprises a specific repeating unit comprising at least one partially defined nucleotide. 4. The method according to claim 3, wherein the specific repeating unit contains degenerate codons. 5. The method according to any one of claims 1 to 4, wherein the two or more proper simple cycles comprise two or more proper simple cycles representing two or more repeating subsequences. 6. The method according to any one of claims 1-5, wherein the sequence graph further comprises two or more alternative paths of two or more alleles. 7. The method of claim 6, wherein two or more alleles comprise an indel or substitution. 8. The method according to claim 6, wherein the substitution comprises a single nucleotide variant (SNV) or a single nucleotide polymorphism (SNP). 9. The method of claim 6, further comprising genotyping two or more alleles using sequence reads aligned to two or more alternative pathways. 10. A system for sequencing genomic loci comprising two or more repeat sequences, wherein the system comprises: a system memory; and one or more processors configured to: (a) obtaining a sequence graph, wherein the sequence graph has a graph data structure with vertices representing nucleotide sequences and directed edges connecting the vertices, and wherein the sequence graph contains two or more proper simple cycles, wherein each proper simple cycle represents a subsequence of repeats; (b) aligning the sequence reads of the sample under study with the reference genome to determine the genomic coordinates of the sequence reads and selecting a subset of the sequence reads; and (c) aligning, using one or more processors, a selected subset of sequence reads with two or more repeat sequences represented by a sequence graph representing the genomic locus.

Images