Computational modeling of loss-of-function based on allele frequency
A computer-based system using probabilistic models enhances cancer diagnosis by distinguishing somatic deletions in cell-free DNA, addressing contamination issues and improving diagnostic accuracy for targeted treatments.
Patent Information
- Application Number
- JP2021549786
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2019-03-25
- Filing Date
- 2020-02-27
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2040-02-27
AI Technical Summary
Existing methods for detecting cancer through analysis of cell-free DNA in bodily fluids face challenges due to low nucleic acid concentrations and contamination with normal DNA, making it difficult to accurately diagnose and characterize the underlying genetic causes of tumors.
A computer-based system uses probabilistic models, such as beta-binomial and binomial distributions, to analyze allele frequencies and distinguish between somatic homozygous and heterozygous deletions in genes like BRCA1/2 by comparing the log-likelihood ratios of competing models, leveraging germline single nucleotide polymorphisms (SNPs) to enhance diagnostic accuracy.
This approach provides highly accurate diagnosis of genetic conditions, enabling precise identification of somatic deletions and informing targeted therapeutic interventions, such as PARP inhibitor administration, based on the genetic analysis of cell-free DNA.
Smart Images

Figure 0007577067000009 
Figure 0007577067000010 
Figure 0007577067000011
Abstract
Description
[Technical field]
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of priority to U.S. Provisional Application No. 62 / 811,159, filed February 27, 2019, and U.S. Provisional Application No. 62 / 823,585, filed March 25, 2019, which are incorporated by reference herein for all purposes. [Background technology]
[0002] background A tumor is an abnormal growth of cells. Fragmented DNA is often released into bodily fluids when cells, such as tumor cells, die. Thus, some of the cell-free DNA in bodily fluids is tumor DNA. Tumors can be benign or malignant. Malignant tumors are often referred to as cancer.
[0003] Cancer is a major cause of disease worldwide. Each year, tens of millions of people worldwide are diagnosed with cancer, and more than half of them will ultimately die from cancer. In many countries, cancer ranks as the second leading cause of death after cardiovascular disease. Early detection is associated with improved outcomes for many cancers.
[0004] Cancer is caused by the accumulation of mutations and / or epigenetic alterations in an individual's normal cells, which at least in part result in inappropriately regulated cell division. Such mutations, or states of genetic material, generally include copy number variations (CNVs), copy number abnormalities (CNAs), single nucleotide variants (SNVs), gene fusions and indels, and epigenetic alterations include modifications to the fifth atom of the six-atom ring of cytosine and its association with chromatin and transcription factors in DNA.
[0005] In certain instances, loss of heterozygosity (LOH) and biallelic copy number loss of homologous recombination repair (HRR) genes (BRCA1 / 2) are associated with loss of tumor suppressor function, resulting in cancer. In many cases, the specific state of a gene of interest can inform the type of treatment. For example, one state of a gene may respond to a set of drugs, while another state of the gene may not. Thus, it is increasingly important to be able to characterize the underlying causes of cancer and other diseases, rather than simply being able to diagnose them.
[0006] Cancer is often detected by tumor biopsy followed by analysis of cells, markers or DNA extracted from the cells. Research is ongoing to detect cancer based on the analysis of bodily fluids. If successful, these tests have the advantage that they are non-invasive and can be performed without identifying suspected cancer cells by biopsy. However, the fact that the amount of nucleic acid in bodily fluids is very low makes it challenging to successfully complete these types of tests. Furthermore, the amount of detectable tumor-associated cell-free nucleic acid in bodily fluids may make the analysis and detection of cancer in cell-free DNA even more difficult. In other words, tumor DNA in bodily fluids may be contaminated with normal DNA, making computational analysis and detection of the specific cause of tumor in cell-free DNA samples difficult. Summary of the Invention [Means for solving the problem]
[0007] Abstract The present disclosure relates to computer techniques that provide accurate diagnosis of various states of genetic material, such as genes sequenced from cell-free DNA in a sample. The states may include, but are not limited to, genetic mutation states, such as somatic homozygous deletion, somatic heterozygous deletion, copy number variation ("CNV") (including wild type, amplification, or loss of a particular copy number), and / or other states. Accurate diagnosis may be based on one or more probability models of these states. For example, a computer system may create competing models that each output the probability that the genetic material is in a particular state.
[0008] Each model can be trained on a set of training samples and output the probability of the genetic material being in each state. For example, a first model can output a first probability that the genetic material contains a somatic homozygous deletion of a particular gene allele. A second model can output a second probability that the genetic material contains a somatic heterozygous deletion of a particular gene allele. Other models can output probabilities of other types of states of the genetic material, such as CNV. The computer system can compare the output of each competing model to determine which model is more likely. For example, the computer system can use the log-likelihood ratio of the competing first and second probabilities to determine whether the genetic material contains a somatic homozygous deletion or a somatic heterozygous deletion.
[0009] In some embodiments, the computer system can use various probability distributions to create the model. For example, the computer system can use beta-binomial distribution, binomial distribution, normal distribution (also referred to as "Gaussian distribution"), and / or other types of probability modeling techniques. The computer system can model the state (such as the number of alleles that support a particular state) based on the training data set to set a baseline expectation of non-normal or tumor state. For example, the computer system can identify the location of germline single nucleotide polymorphisms (SNPs) observed in "normal" or non-tumor samples, e.g., samples in which no somatic variants are observed. These samples will also be referred to as tumor not detected (TND) samples.
[0010] Since the TND sample is normal, the computer system can assume that the germline SNP position is not the cause of the non-normal condition. In this way, the computer system can utilize these SNP sites as reference expected values for modeling the allele count for the probabilistic determination of this condition. For example, deviations from the observed nucleotide calls at each SNP position can indicate the probability that such deviations are the cause of a particular condition, such as a tumor or other non-normal condition. Thus, the computer system can train a model based on the expected values derived from the calculation results on the data from the germline SNPs of the TND sample. Such calculated data can include, for each SNP site: heterozygosity incidence, standard deviation of MAF, genotype, germline incidence (priori), and / or other data that can inform the analysis of the individual's sample.
[0011] Using the calculated expected value, the computer system can model the state based on the sequence read data of the sample of the individual to be tested that aligns to the region of interest, for example, the region upstream, downstream, and including the gene of interest. In some embodiments, the sequence read data of the molecule obtained from the sample of the individual can be aligned to a reference genome to identify the allele (mutant or wild type) that the original molecule supports. Based on the alignment of the sequence read data obtained from the sample of the individual, the computer system can identify the number of molecules that support alternative alleles and calculate the total number of molecules. The computer system can model these and / or other data derived from the sample of the individual using the expected value data calculated from each germline SNP in the region of interest. In some examples, the sequencing can be based on targeted sequencing of plasma cell-free DNA (cfDNA).
[0012] In one aspect, the present disclosure relates to an improved computer system for distinguishing between somatic homozygous deletion and somatic heterozygous deletion of a gene in a sample that does not exhibit a germline deletion of the gene. The computer system may include a processor programmed to create a first model of allele counts representing somatic homozygous deletion based on the location of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene according to a first probability distribution. The processor may further create a second model of allele counts in the sample representing somatic heterozygous deletion based on the location of one or more germline SNPs according to a second probability distribution. The processor may compare a first output of the first model with a second output of the second model. The processor may create a prediction that a somatic homozygous deletion for the gene is present in the sample based on the comparison.
[0013] In some embodiments, the first model can represent a first probability that the sample contains a somatic homozygous deletion, and the second model represents a second probability that the sample contains a somatic heterozygous deletion.
[0014] In some embodiments, the first probability distribution is the same type of probability distribution as the second probability distribution.
[0015] In some embodiments, to create the first model, the processor is programmed to determine one or more parameters for inputs to the first probability distribution.
[0016] In some embodiments, the first probability distribution comprises a beta-binomial distribution, a binomial distribution, or a normal distribution.
[0017] In some embodiments for creating the first model of allele count, the processor may further determine a prevalence of heterozygosity of one or more germline SNPs in the training set of samples for input to the first probability distribution.
[0018] In some embodiments, the training set of samples may include a number of tumor not detected (TND) samples.
[0019] In some embodiments, to create the first model of allele counts, the processor may further determine a standard deviation of minor allele frequencies (MAFs) associated with each of the one or more germline SNPs in the training set of samples for input to the first probability distribution.
[0020] In some embodiments, to generate the first model, the processor may further determine a number of molecules in the sample that support the variant allele for input to the first probability distribution.
[0021] In some embodiments, to generate the first model, the processor may further determine a total number of molecules in the sample for input to the first probability distribution.
[0022] In some embodiments, to create the first model, the processor may further calculate a first likelihood of the allele count at one or more germline SNP positions in the sample assuming a somatic homozygous deletion based on sequence read data coverage associated with a somatic homozygous deletion.
[0023] In some embodiments, to generate the second model, the processor may further calculate a second likelihood of the allele count at one or more germline SNP positions in the sample assuming a somatic heterozygous deletion based on sequence read data coverage associated with a somatic heterozygous deletion.
[0024] In some embodiments, to generate the second model, the processor may further determine a mean tumor fraction estimated from the samples for input to a second probability distribution for the second model.
[0025] In some embodiments, tumor content may be estimated based on sequence coverage information.
[0026] In some embodiments, to generate the second model, the processor may further determine a standard deviation of the tumor content estimated from the sample for input to a second probability distribution for the second model.
[0027] In some embodiments, the processor is further capable of accessing the plurality of samples, identifying a set of samples from the plurality of samples that contain a germline deletion, filtering and outputting the set of samples from the plurality of samples, and identifying the presence of a somatic homozygous deletion or a somatic heterozygous deletion from the filtered plurality of samples.
[0028] In some embodiments, the first output may include a first probability for the presence of a somatic homozygous deletion and the second output may include a second probability for the presence of a somatic heterozygous deletion.
[0029] In some embodiments, to compare the first output of the first model and the second output of the second model, the processor may further perform a log-likelihood function based on the first output and the second output.
[0030] In some embodiments, the gene may include BRCA1, BRCA2, or ATM.
[0031] In another aspect, the present disclosure relates to a system.The system may include a processor programmed to: generate a first probability that a gene in a sample contains a somatic homozygous deletion; generate a second probability that a gene in a sample contains a somatic heterozygous deletion; compare the first probability with the second probability; and generate a prediction of whether the sample contains a somatic homozygous deletion or a somatic heterozygous deletion.
[0032] In another aspect, the disclosure relates to a system that may include a processor programmed to generate a first probability that genetic material in a sample comprises a first state, generate a second probability that genetic material in the sample comprises a second state, compare the first probability to the second probability, and generate a prediction of whether the sample comprises the first state or the second state.
[0033] In some embodiments, the first condition comprises a somatic homozygous deletion and the second condition comprises a somatic heterozygous deletion.
[0034] In some embodiments, the first condition may include a first copy number variant (CNV) and the second condition may include a second CNV that is different from the first CNV.
[0035] In some embodiments, the first CNV and / or the second CNV may be associated with a deleterious condition.
[0036] In some embodiments, to generate the first probability, the processor may further access the locations of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene and determine a standard deviation of minor allele frequencies (MAFs) associated with each of the one or more germline SNPs in the training set of samples.
[0037] In some embodiments, to generate the first probability, the processor may further determine a standard deviation of minor allele frequencies (MAFs) associated with each of the one or more germline SNPs in the training set of samples for input to the probability distribution.
[0038] In another aspect, the disclosure relates to a method implemented by a processor. The method may include creating, by the processor, a first model of allele counts representing somatic homozygous deletions based on the positions of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene according to a first probability distribution. The method may further include creating, by the processor, a second model of allele counts in the sample representing somatic heterozygous deletions based on the positions of one or more germline SNPs according to a second probability distribution. The method may include comparing, by the processor, a first output of the first model and a second output of the second model. The method may further include creating, by the processor, a prediction that a somatic homozygous deletion for the gene is present in the sample based on the comparison.
[0039] In another aspect, the present disclosure relates to another method implemented by a processor. The method may include: generating, by the processor, a first probability that a gene in a sample contains a somatic homozygous deletion. The method may further include: generating, by the processor, a second probability that a gene in a sample contains a somatic heterozygous deletion. The method may further include: comparing, by the processor, the first probability with the second probability. The method may further include: generating, by the processor, a prediction of whether the sample contains a somatic homozygous deletion or a somatic heterozygous deletion.
[0040] In another aspect, the present disclosure relates to another processor-implemented method.
[0041] The method may include generating, by the processor, a first probability that the genetic material in the sample comprises a first state. The method may further include generating, by the processor, a second probability that the genetic material in the sample comprises a second state. The method may further include comparing, by the processor, the first probability and the second probability. The method may further include generating, by the processor, a prediction of whether the sample comprises the first state or the second state.
[0042] In another aspect, the disclosure relates to a method for administering to a subject determined to have a somatic homozygous deletion based on the disclosure herein, a therapeutic intervention effective to treat a cancer associated with the somatic homozygous deletion.
[0043] In some embodiments, the therapeutic intervention may include a poly ADP ribose polymerase (PARP) inhibitor. Examples of PARP inhibitors include OLAPARIB, TALAZOPARIB, RUCAPARIB, NIRAPARIB (trade name ZEJULA), among others.
[0044] In some embodiments, the therapeutic intervention may include a base excision repair (BER) inhibitor. For example, OLAPARIB may inhibit BER.
[0045] In another aspect, the present disclosure relates to a method for administering to a subject determined to have a particular state of genetic material based on the disclosure herein, a therapeutic intervention effective to treat a disease associated with the state of the genetic material.
[0046] In another aspect, the disclosure relates to a method for administering a therapeutic intervention to exclude a PARP inhibitor to a subject determined to not have a somatic homozygous deletion based on the disclosure herein.
[0047] In some embodiments of all aspects of the present disclosure, the results of the system and / or method disclosed herein are used as input to generate a report. The report may be in paper or electronic format. For example, information about the deletion or other condition of genes and / or genetic material determined by the method or system disclosed herein and / or information derived therefrom can be shown in such a report. The method or system disclosed herein may further include communicating the report to a third party, such as the subject from whom the sample is derived or a medical professional. The various operations of the methods disclosed herein or performed by the systems disclosed herein may be performed at the same time or at different times and / or in the same geographic location or different geographic locations, e.g., countries. The various steps of the methods disclosed herein may be performed by the same person or different people. [Brief description of the drawings]
[0048] [Figure 1] FIG. 1 illustrates an example of a system for training a model to predict a state of genetic material based on the probability of each state according to an embodiment of the present disclosure.
[0049] [Diagram 2] FIG. 2 illustrates a schematic of determining allele numbers for germline SNPs to predict genetic status according to an embodiment of the present disclosure.
[0050] [Diagram 3] FIG. 3 illustrates a process of predicting somatic homozygous or heterozygous deletions based on a trained model according to an embodiment of the present disclosure.
[0051] [Figure 4] FIG. 4 illustrates a process of predicting a genetic material state based on a trained model according to an embodiment of the present disclosure.
[0052] [Diagram 5] FIG. 5 illustrates types of somatic deletions according to an embodiment of the present disclosure.
[0053] [Figure 6A] FIG. 6A illustrates an exemplary plot of a BRCA1 homozygous deletion according to an embodiment of the present disclosure.
[0054] [Figure 6B] FIG. 6B illustrates an exemplary plot of a BRCA2 heterozygous deletion according to an embodiment of the present disclosure.
[0055] [Figure 7A] FIG. 7A illustrates an exemplary plot of het occurrence in TND samples according to an embodiment of the present disclosure.
[0056] [Figure 7B] FIG. 7B illustrates an exemplary plot of MAF across a TND sample, according to an embodiment of the present disclosure.
[0057] [Figure 8A] FIG. 8A illustrates an exemplary plot of MAF values for BRCA1 according to an embodiment of the present disclosure.
[0058] [Figure 8B] FIG. 8B illustrates an exemplary plot of MAF values for BRCA2 according to an embodiment of the present disclosure.
[0059] [Figure 9A] FIG. 9A illustrates an exemplary plot of a score comparison of a beta-binomial model and a binomial model for a BRCA2 panel according to an embodiment of the present disclosure.
[0060] [Figure 9B] FIG. 9B illustrates an exemplary plot of a score comparison of a beta-binomial model and a Gaussian distribution model for a BRCA2 panel according to an embodiment of the present disclosure.
[0061] [Figure 10A] FIG. 10A illustrates an exemplary plot of LLR score distribution for BRCA1 negative samples according to an embodiment of the present disclosure.
[0062] [Figure 10B] FIG. 10B illustrates an exemplary plot of LLR score distribution for BRCA2 negative samples according to an embodiment of the present disclosure.
[0063] [Figure 11A] FIG. 11A illustrates an exemplary plot of the loss of limit of detection (LoD) for BRCA1 according to an embodiment of the present disclosure.
[0064] [Figure 11B] FIG. 11B illustrates an exemplary plot of HRR homozygous deletion of LoD for BRCA1 according to an embodiment of the present disclosure.
[0065] [Figure 12A] FIG. 12A illustrates an exemplary plot of loss of LoD for BRCA2 according to an embodiment of the present disclosure.
[0066] [Figure 12B] FIG. 12B illustrates an exemplary plot of HRR homozygous deletion of LoD for BRCA2 according to an embodiment of the present disclosure.
[0067] [Figure 13] FIG. 13 illustrates an exemplary plot of TF prevalence versus cancer type, according to an embodiment of the present disclosure.
[0068] [Figure 14] FIG. 14 illustrates an exemplary plot of density of LLR scores for BRCA1 and BRCA2 according to an embodiment of the present disclosure.
[0069] [Figure 15] FIG. 15 illustrates an exemplary chart of the prevalence of BRCA2 homozygous deletions according to an embodiment of the present disclosure.
[0070] [Figure 16] FIG. 16 illustrates an exemplary chart of the prevalence of BRCA1 homozygous deletions according to an embodiment of the present disclosure.
[0071] [Figure 17] FIG. 17 illustrates an example of a homozygous deletion of BRCA2 and potential clinical actionability according to an embodiment of the present disclosure.
[0072] [Figure 18A] FIG. 18A illustrates an exemplary plot of a BRCA1 homozygous deletion according to an embodiment of the present disclosure.
[0073] [Figure 18B] FIG. 18B illustrates an exemplary plot of a BRCA1 homozygous deletion according to an embodiment of the present disclosure.
[0074] [Figure 19A] FIG. 19A illustrates an exemplary plot of a BRCA2 homozygous deletion according to an embodiment of the present disclosure.
[0075] [Figure 19B] FIG. 19B illustrates an exemplary plot of a BRCA2 homozygous deletion according to an embodiment of the present disclosure.
[0076] [Figure 20A] FIG. 20A illustrates an exemplary plot of BRCA1 biallelic somatic copy number loss according to an embodiment of the present disclosure.
[0077] [Figure 20B] FIG. 20B illustrates an exemplary plot of BRCA1 LOH according to an embodiment of the present disclosure.
[0078] [Figure 21A] FIG. 21A illustrates an exemplary plot of BRCA2 biallelic somatic copy number loss according to an embodiment of the present disclosure.
[0079] [Figure 21B] FIG. 21B illustrates an exemplary plot of BRCA2 LOH according to an embodiment of the present disclosure.
[0080] [Figure 22] FIG. 22 illustrates a plot of the incidence of BRCA1 and BRCA2 somatic deletions according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0081] definition A subject refers to an animal, such as a mammalian species (preferably human) or an avian (e.g., avian) species, or other organism, such as a plant. More specifically, a subject may be a vertebrate, such as a mammal, such as a mouse, a primate, an ape, or a human. Animals include domestic animals, game animals, and pet animals. A subject may be a healthy individual, an individual suspected of having symptoms or signs or having a disease or a predisposition to a disease, or an individual in need of treatment or suspected of needing treatment.
[0082] Genetic variant refers to an alteration, variant or polymorphism in a nucleic acid sample or genome of a subject. Such alteration, variant or polymorphism may be with respect to a reference genome, which may be the reference genome of a species (e.g., in humans, hG19 or hG38), a subject or other individual. Mutations include one or more single nucleotide variations (SNVs), insertions, deletions, repeats, small insertions, small deletions, small repeats, structural variant junctions, variable length tandem repeats, and / or flanking sequences, and copy number variants (CNVs), transversions, gene fusions and other rearrangements are also forms of genetic variation. Mutations may be base changes, insertions, deletions, repeats, copy number variations, transversions, or combinations thereof.
[0083] Cancer markers are genetic variants associated with the presence of cancer or the risk of developing cancer.Cancer markers can provide an indication that a subject has a higher risk of developing cancer than the same age- and sex-matched subjects who have cancer or do not have cancer markers.Cancer markers may or may not be the cause of cancer.
[0084] As used herein, "nucleic acid tag" refers to a short nucleic acid (e.g., less than about 500 nucleotides long, about 100 nucleotides long, about 50 nucleotides long, or about 10 nucleotides long) used to distinguish nucleic acids from different samples of different types or that have been subjected to different processes (e.g., representing sample indexes), or from different nucleic acid molecules in the same sample (e.g., representing molecular barcodes). Nucleic acid tags include a predetermined, fixed, non-random, random, or semi-random oligonucleotide sequence. Such nucleic acid tags can be used to label different nucleic acid molecules or different nucleic acid samples or subsamples. Nucleic acid tags may be single-stranded, double-stranded, or at least partially double-stranded. Nucleic acid tags have the same length or varying lengths, as appropriate. Nucleic acid tags may include double-stranded molecules with one or more blunt ends, may include 5' or 3' single-stranded regions (e.g., overhangs), and / or may include one or more other single-stranded regions elsewhere within a given molecule. Nucleic acid tags can be attached to one or both ends of other nucleic acids (e.g., sample nucleic acids to be amplified and / or sequenced). Nucleic acid tags can be decoded to reveal information such as the sample of origin, form, or processing of a given nucleic acid. For example, nucleic acid tags can also be used to enable pooling and / or parallel processing of multiple samples containing nucleic acids carrying different molecular barcodes and / or sample indexes, where the nucleic acids are then deconvoluted by detecting (e.g., reading) the nucleic acid tags. Nucleic acid tags can also be referred to as identifiers (e.g., molecular identifiers, sample identifiers). Additionally or alternatively, nucleic acid tags can be used as molecular identifiers (e.g., to distinguish between amplicons of different molecules or different parent molecules in the same sample or subsample). This includes, for example, uniquely tagging different nucleic acid molecules in a given sample, or non-uniquely tagging such molecules.In cases where non-unique tagging is applied, a limited number of tags (i.e., molecular barcodes) can be used to tag each nucleic acid molecule such that different molecules can be identified based on their endogenous sequence information (e.g., start and / or stop positions, subsequences at one or both ends of the sequence, and / or length of the sequence) in combination with at least one molecular barcode. Typically, a sufficient number of different molecular barcodes are used such that the probability that any two molecules may have the same endogenous sequence information (e.g., start and / or stop positions, subsequences at one or both ends of the sequence, and / or length) and also have the same molecular barcode is low (e.g., less than about 10%, less than about 5%, less than about 1%, or less than about 0.1% chance).
[0085] Adapters are short nucleic acids (e.g., less than 500, 100, or 50 nucleotides long) that are typically at least partially double-stranded for ligation to either or both ends of a sample nucleic acid molecule (one adapter at each end). Adapters can include primer binding sites that allow amplification of the nucleic acid molecule at both ends adjacent to the adapter, and / or sequencing primer binding sites that include primer binding sites for next-generation sequencing (NGS). Adapters can also include binding sites for capture probes, e.g., oligonucleotides bound to the support of a flow cell. Adapters can also include barcodes as described above. The barcodes are preferably positioned relative to the primers and sequencing primer binding sites such that the barcodes are included in the amplicon and sequencing read data of the nucleic acid molecule. Adapters of the same or different sequences can be ligated to each end of the nucleic acid molecule. The same adapters are often ligated to each end, except that the barcodes are different. A preferred adaptor is a Y-shaped adaptor, one end of which is blunt-ended or tailed as described herein for joining to a nucleic acid molecule (which is also blunt-ended or tailed with one or more complementary nucleotides), and the other end of the Y-shaped adaptor comprises a non-complementary sequence that does not hybridize to form a double strand. Another preferred adaptor is a bell-shaped adaptor, also having a blunt or tailed end for joining to the nucleic acid to be analyzed.
[0086] As used herein, the term "sequencing" refers to any of several techniques used to determine the sequence of a biomolecule, e.g., a nucleic acid such as DNA or RNA. Exemplary sequencing methods include, but are not limited to, targeted sequencing, single molecule real-time sequencing, exon sequencing, electron microscope-based sequencing, panel sequencing, transistor-mediated sequencing, direct sequencing, random shotgun sequencing, Sanger dideoxytermination sequencing, whole genome sequencing, sequencing by hybridization, pyrosequencing, capillary electrophoresis, duplex sequencing, cycle sequencing, single base extension sequencing, solid-phase sequencing, high-throughput sequencing, massively parallel signature sequencing, emulsion PCR, co-amplification in low denaturation temperature-PCR (COLD-PCR), multiplex PCR, reversible dye terminator sequencing, paired-end sequencing, near-term sequencing, exonuclease sequencing, sequencing by ligation, short-read sequencing, single molecule sequencing, sequencing by synthesis, real-time sequencing, reverse terminator sequencing, nanopore sequencing, 454 sequencing, Solexa Genome Examples of sequencing methods include Analyzer sequencing, SOLiD™ sequencing, MS-PET sequencing, and combinations thereof. In some embodiments, sequencing can be performed by a genetic analyzer, such as a genetic analyzer commercially available from Illumina or Applied Biosystems.
[0087] The phrase "next-generation sequencing" or NGS refers to a sequencing technology that has increased throughput, for example, compared to classical Sanger and capillary electrophoresis-based approaches, with the ability to generate hundreds of thousands of relatively small sequence read data at a time. Some examples of next-generation sequencing techniques include, but are not limited to, sequencing by synthesis, sequencing by ligation, and sequencing by hybridization.
[0088] DNA (deoxyribonucleic acid) is a nucleotide chain that contains four types of nucleotides based on adenine (A), thymine (T), cytosine (C), and guanine (G). RNA (ribonucleic acid) is a nucleotide chain that contains four types of nucleotides based on A, uracil (U), G, and C. Certain pairs of nucleotides bind specifically to each other in a complementary manner (called complementary base pairing). In DNA, adenine (A) pairs with thymine (T) and cytosine (C) pairs with guanine (G). In RNA, adenine (A) pairs with uracil (U) and cytosine (C) pairs with guanine (G). When a first nucleic acid strand binds to a second nucleic acid strand that is composed of nucleotides that are complementary to the nucleotides in the first strand, the two strands combine to form a duplex. As used herein, "nucleic acid sequencing data," "nucleic acid sequencing information," "nucleic acid sequence," "nucleotide sequence," "genomic sequence," "gene sequence," or "fragment sequence," or "nucleic acid sequencing read data" refers to any information or data that indicates the order of nucleotide bases (e.g., adenine, guanine, cytosine, and thymine or uracil) in a molecule of nucleic acid such as DNA or RNA (e.g., a whole genome, a whole transcriptome, an exome, an oligonucleotide, a polynucleotide, or a fragment). It should be understood that the teachings of the present invention contemplate sequence information obtained using all available techniques, platforms, or technologies, including, but not limited to: capillary electrophoresis, microarrays, ligation-based systems, polymerase-based systems, hybridization-based systems, direct or indirect nucleotide specification systems, pyrosequencing, ion- or pH-based detection systems, and electronic signature-based systems.
[0089] A "polynucleotide," "nucleic acid," "nucleic acid molecule," or "oligonucleotide" refers to a linear polymer of nucleosides (including deoxyribonucleosides, ribonucleosides, or analogs thereof) joined by internucleoside linkages. Typically, a polynucleotide contains at least three nucleosides. Oligonucleotides often range in size from a few monomeric units, e.g., 3-4, to hundreds of monomeric units. Whenever a polynucleotide is represented by a sequence of letters, such as "ATGCCTG," it will be understood that the nucleotides are in 5' to 3' order from left to right, with "A" indicating deoxyadenosine, "C" indicating deoxycytidine, "G" indicating deoxyguanosine, and "T" indicating thymidine, unless otherwise noted. The letters A, C, G, and T can be used to refer to the bases themselves, nucleosides, or nucleotides containing the bases, as is standard in the art.
[0090] The phrase "sequence read data coverage" refers to the number of sequence reads that align to a locus of a reference sequence. "Sequence coverage information" refers to information that conveys the sequence read data coverage of a given locus of a reference sequence. Sequence coverage information can include the number or identity of sequence reads that align to a locus and / or other information that indicates the sequence read data coverage at the locus.
[0091] The phrase "molecular coverage" refers to the number of molecules that cover the locus of a reference sequence.Molecules can be identified based on sequence read data and molecular barcodes as described herein.In this way, molecules can be determined to cover the locus of a reference sequence based on sequence read data generated from molecules that align to the locus.
[0092] Reference sequence is a known sequence that is used to compare with experimentally determined sequence.For example, known sequence can be the whole genome, chromosome, or any segment thereof.Reference typically comprises at least 20;50;100;200;250;300;350;400;450;500;1,000;10,000;100,000;1,000,000;10,000,000;100,000,000;1,000,000,000 or more nucleotides.Reference sequence can be aligned with a single continuous sequence of genome or chromosome, or can comprise non-contiguous segments that align with different regions of genome or chromosome.Reference human genome includes, for example, hG19 and hG38.
[0093] The term "designated position" in a reference sequence refers to a genomic coordinate in the reference sequence.
[0094] A first single-stranded nucleic acid sequence overlaps with a second single-stranded sequence when the first nucleic acid sequence or its complement and the second nucleic acid sequence or its complement align with overlapping but non-identical segments of a contiguous reference sequence, such as the sequences of human chromosomes. A fully or partially double-stranded nucleic acid overlaps with another fully or partially double-stranded nucleic acid when either of its strands overlaps with the strand of the other nucleic acid.
[0095] A "C" to "T" variant or transversion refers to the presence of a base "T" in a sequenced polynucleotide at a coordinate position occupied by a base "C" in a reference sequence. A "G" to "A" variant or transversion refers to the presence of a base "A" in a sequenced polynucleotide at a coordinate position occupied by a base "G" in a reference sequence.
[0096] A nucleic acid molecule can be conceptually divided into a 5'-end, an internal portion and a 3'-end. The terminal end can be designated based on a certain number of nucleotides from the end. For example, the 5'-end is represented by, for example, 20 terminal nucleotides to the 5'-end. The 3'-end is represented by, for example, 20 terminal nucleotides to the 3'-end. Alternatively, a nucleic acid molecule can be divided into an end portion and a remaining portion as described.
[0097] The term "minor allele frequency" ("MAF") refers to the frequency with which a minor allele (e.g., the least common allele) occurs in a given population of nucleic acids, such as a sample.
[0098] "Tumor content" (TF) refers to the proportion of DNA molecules associated with tumors in a given sample. TF can be derived based on detecting reduced coverage of variant alleles in tumor cells. Lower TF in a given sample can affect the MAF of a given variant allele in a given sample, and therefore the detectability of the given variant allele.
[0099] The term "tumor non-detected" or "TND" refers to samples in which no somatic single nucleotide variants, insertion-deletions, copy number variants, or fusions were detected.
[0100] The terms "processing", "calculating" and "comparing" can be used interchangeably. This term can refer to determining differences, such as differences in numbers or sequences. For example, gene expression, copy number variation (CNV), indels, and / or single nucleotide variants (SNV) values or sequences can be processed.
[0101] Adapters are artificially synthesized sequences that can be linked to a nucleic acid molecule or polynucleotide sequence by any approach, including ligation, hybridization, and / or amplification. Adapters are usually short nucleic acids (e.g., less than 500, 100, or 50 nucleotides long) that are at least partially double-stranded for linking to either or both ends of a sample nucleic acid molecule. Adapters can include primer binding sites that allow amplification of the nucleic acid molecule flanked by the adapter at both ends, and / or sequencing primer binding sites that include primer binding sites for next-generation sequencing (NGS). Adapters can also include binding sites for capture probes, such as oligonucleotides bound to the support of a flow cell. Adapters can also include barcodes as described above. Tags are preferably positions relative to the primers and sequencing primer binding sites so that the tags are included in the amplicon and sequencing read data of the nucleic acid molecule. The same or different adapters can be linked to each end of a nucleic acid molecule. Often the same adapter is linked to each end, except that the tags are different. A preferred adaptor is a Y-shaped adaptor, one end of which is blunt or tailed as described herein, for joining to a nucleic acid molecule (which is also blunt or tailed with one or more complementary nucleotides). Another preferred adaptor is a bell-shaped adaptor, also having a blunt or tailed end for joining to a nucleic acid to be analyzed. Detailed Description
[0102] 1 illustrates an example of a system 100 for training and using computer models to predict the state of genetic material based on the probability of each state according to an embodiment of the present disclosure. The system can process a sample 101 to train one or more models 140 (illustrated as models 140a...n), each of which outputs the probability that genetic material, such as a sample from an individual, is in a particular state under test (IUT) 111. In some examples, the sample 101 can include a panel of different genes of interest to be studied.
[0103] For example, the system can use model 140a to determine the probability that a sample from IUT111 contains a somatic homozygous deletion associated with a gene. The system can use another model 140b to determine the probability that a sample from IUT111 contains a somatic heterozygous deletion associated with a gene. The system can then compare the probabilities with each other to determine whether a somatic homozygous deletion or a somatic heterozygous deletion is more likely. The system can similarly provide other types of highly accurate diagnoses based on competing probabilities. For example, the system can model CNVs by modeling the probabilities of different copy numbers. Based on a comparison of the output probabilities of each model (which may each correspond to a different copy number prediction), the system can determine the CNVs in the sample of IUT111.
[0104] System 100 may include a sequencing system 102, a computer system 110, and / or other components. Note that sequencing system 102 and computer system 110 may be isolated from each other or connected to each other via a computer network (not illustrated). Sequencing system 102 may include a sample collection and preparation pipeline 103, a sequencing pipeline 105, and a sequencing read data storage 109, and / or other components. Sequencing pipeline 105 may include one or more sequencing devices 107 (illustrated in FIG. 1 as sequencing devices 107a...n).
[0105] The computer system 110 may include a sequence analysis pipeline 112, a processor 120, a storage device 122, a data pre-processing subsystem 124, a classifier 130, a model validator 132, and / or other components.
[0106] The sequence analysis pipeline 112 may include a sequence quality control (QC) component 113, an alignment component 114, an other analysis component 115, and an analysis QC component 116. Output from the sequence analysis pipeline 112 may be stored in an analysis data store 117. A data pre-processing subsystem 124 may pre-process data from the sequence analysis pipeline 112 to create a training data set 125. For example, the training data set 125 may include tumor no detected data ("TND") (if cancer is diagnosed) or otherwise normal samples (if other types of diseases or conditions are diagnosed) from among the samples 101. Examples disclosed throughout may refer to TND samples for illustration purposes.
[0107] In some embodiments, the training dataset 125 may be stored in the training data storage 126. To illustrate an exemplary operation of the processor 120, reference will be made to FIG. 2. FIG. 2 illustrates a schematic diagram 200 of determining allele counts for germline SNPs to predict the status of a gene 201 according to an embodiment of the present disclosure. In some examples, the processor 120 may identify germline SNPs of TND samples for a region of interest 201 around the gene 201. In one training example, germline SNPs were selected from 28,199 samples. Of these samples, 5105 samples (18%) were identified as having TND and were used for the population allele / genotype frequencies. Germline SNPs were selected that met the following criteria: (1) were within 3 Mb of the selected gene (BRCA1, BRCA2, ATM, etc.), (2) the frequency of heterozygosity calls (MAF>25% and MAF<75%) across 5105 TND samples was between 5% and 95%, and (3) the variant was not called somatically in all 28,199 samples. The region of interest 203 can include N bases upstream from the start of the gene 201 and M bases downstream from the end of the gene 201. The values of N and M can be the same or different. In some examples, N and M can each be 3,000,000 nucleotides (3 Mb).
[0108] In the illustrated example of FIG. 2, the reference wild type nucleotide at SNP site (i) (illustrated as SNP(i)) may be "G". Across the TND samples, the nucleotides called at that position may differ from one another. Because the TND sample is normal, processor 120 may assume that SNP(i) and other SNP sites in the TND sample are not causative of a tumor or non-normal condition. In this manner, these SNP sites may each serve as a reference expectation value for modeling allele counts for probabilistic determination of the state of the gene. For example, deviations from the observed nucleotide calls at each SNP site may indicate the probability that such deviations are causative of a particular condition, such as a tumor or other non-normal condition of gene 201. Thus, processor 120 may train model 140 based on expectation values derived from calculations on data from germline SNPs of the TND samples. Such calculated data may include, for each SNP site, heterozygosity rates, standard deviations of minor allele frequencies (MAFs), genotypes, germline rates (prior), and / or other data.
[0109] Using the calculated expectation value, the processor 120 can model the state of the gene 201 based on sequence read data of the IUT 111 samples that align to the region of interest 203. For example, the processor 120 can create competitive models 140 that each output a score that represents the probability that the gene 201 is in a particular state. The processor 120 can compare the scores to calculate a predicted score, which can be compared to a threshold score to determine the state of the gene 201. The processor 120 can calculate the threshold score based on observed data from training samples, as described further below.
[0110] In some embodiments, sequence read data of molecules generated from IUT111 samples can be aligned to a reference genome to identify the allele (mutant or wild type) supported by the original molecule. IUT111 samples can be prepared in sample collection and preparation pipeline 103 and sequenced in sequencing pipeline 105. Each molecule can be associated with sequence read data. The number of sequencing read data from the number of molecules of the sample from IUT111 can cover a given germline SNP site.
[0111] Based on the alignment of sequence read data generated from the IUT111 samples, the processor 120 can identify the number of molecules supporting the SNP alleles and calculate the total number of molecules. The processor 120 can model these and / or other data from the IUT111 samples with expected value data calculated from each of the germline SNPs in the region of interest 203. For example, the processor 120 can generate a first output of the allele count model 140a representing the probability of a first state of the gene 201 and a second output of the allele count model 140b representing the probability of a second state of the gene 201.
[0112] The processor 120 can implement different types of probability distributions to create the model 140. Furthermore, the model 140 can model different types of states of the genes 201, or more generally, different types of states of genetic material. Attention is now directed to examples of modeling and types of states modeled by the processor 120.
[0113] Generally speaking, the processor 120 may implement (be programmed by) the classifier 130. Alternatively, it should be noted that the classifier 130 may include a hardware module. In either event, the classifier 130 (which may program the processor 120) may model the state of a gene (such as gene 201 illustrated in FIG. 2) based on alleles detected in a region of interest associated with the gene. More specifically, based on the training dataset 125, the classifier 130 may determine one or more probability models 140 (illustrated as models 140a, 140b, . . . , 140n) for a particular state of the gene based on the location of germline single nucleotide polymorphisms (SNPs) in the region of interest (such as region of interest 203 illustrated in FIG. 2). Each model 140 may correspond to a respective probability of the state of the gene. The location of the SNPs may be based on sequencing read data generated by the sequencing system 102 from various samples 101. The condition can include, but is not limited to, the mutational state of a gene, such as somatic homozygous deletion, somatic heterozygous deletion, copy number variation ("CNV") (including wild type, gain, or loss of a particular copy number), and / or other condition of a gene.
[0114] In various embodiments, the classifier 130 can apply a model, which may be created based on the training data set 125, to determine the state of genes in a sample from an individual. For example, the classifier 130 can determine the probability that genes in a sample of cfDNA from an individual contain somatic homozygous deletions, somatic heterozygous deletions, and / or other conditions that may correlate with a disease, such as cancer, or other health conditions. Based on the prediction, a precision treatment can be tailored to the individual. In this manner, the computer system 110 can be improved to provide advanced diagnostic capabilities based on non-invasive analysis of genetic material, such as cfDNA.
[0115] It should be noted that although the examples described herein may relate to determining genetic status, the status of other genetic material, such as chromosomes, exomes, and / or other genetic material, may be determined as well. For example, CNVs may be determined for chromosomes, exomes, and / or other genetic material. Having provided an explanation of the functionality of the classifier 130, attention is now directed to a more detailed example of determining genetic status by training various models 140 and using the models 140 to predict the probability that a particular sample under test will exhibit a particular genetic status.
[0116] Training the model based on TND samples
[0117] In some embodiments, the classifier 130 uses data from the samples 101. The data may include a set of tumor-free samples ("TND" samples). The classifier 130 may use the TND samples to determine the heterozygosity rate of each germline SNP in the TND samples and the standard deviation of the minor allele frequency (MAF) of each germline SNP in the TND samples. Note that in the formulas and calculations described throughout this disclosure, variance may be used instead of standard deviation, as long as appropriate adjustments are made for such calculations that use variance instead of standard deviation. The heterozygosity rate and standard deviation may provide a baseline expectation for "normal" samples, i.e., samples that do not exhibit a disease state. The classifier 130 may use the germline rate (previous) g for each site i. i An exemplary calculation of the incidence of heterozygosity for each germline SNP can be given by equation (1):
number
[0118] Modeling somatic homozygous deletions
[0119] The classifier 130 can create a first model 140a for the allele count based on the location of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene according to a probability distribution. The first model can represent (e.g., model) somatic homozygous deletion of the gene. For example, given the heterozygous occurrence rate (Equation (1)) of each germline SNP in the TND samples and the standard deviation of the MAF of each germline SNP in the TND samples, the classifier 130 can model the probability that a gene in a particular sample from an individual is associated with a somatic homozygous deletion. To do so, the classifier 130 can access the number of molecules supporting the presence of a somatic homozygous deletion in the gene in the IUT111 sample and the total number of molecules in the IUT111 sample. For example, the classifier 130 can create a model, such as model 140a, that represents the probability that a gene in a sample has a somatic homozygous deletion. In some embodiments, the classifier 130 may use a beta-binomial probability distribution to create the model 140a, although other probability distributions, such as a binomial probability distribution, a normal distribution (Gaussian distribution), and / or other probability modeling, may also be used.
[0120] The beta-binomial distribution is a binomial distribution with n Bernoulli trials where the probability of success on each trial is fixed but randomly drawn from the beta distribution. The beta-binomial distribution can use two parameters: α and β (uniquely determined by the mean / standard deviation of the distribution). When n=1, the distribution reduces to a Bernoulli distribution. With α=β=1, it is a separate uniform distribution from 0 to n.
[0121] The binomial distribution is the probability distribution of a binomial random variable. A binomial random variable is the number of successes in N repeated trials of a binomial experiment. The binomial distribution has the following properties: The mean of the distribution (μx) is n* equal to P; variance is n * P * (1-P); and the standard deviation (σx) is given by equation (2):
number
[0122] The normal distribution may be defined by the normal equation:
number
[0123] An example of applying the beta-binomial probability distribution will now be described for illustrative purposes. Those skilled in the art will appreciate that binomial, normal, and / or other probability distributions can be similarly used based on the disclosure herein. For the beta-binomial probability distribution, the classifier 130 can use the dbetabinom function in the VGAM package of the R Project according to equation (4):
number
[0124] The classifier 130 may generate a first probability output (L1) of the first model 140a according to equation (5):
number
[0125] Modeling somatic heterozygous deletions
[0126] The classifier 130 can create a second model 140b of the allele count in the sample based on the location of one or more germline SNPs with a probability distribution. The second model 140b can represent (e.g., model) somatic heterozygous deletions of genes. Because detection of heterozygous deletions can be influenced by TFs, the classifier 130 can determine the mean, mu.tf (which may be denoted as μ.tf), and standard deviation, sd.tf (which may be denoted as σ.tf), of the TFs based on the coverage of the read data in the IUT 111 sample (sequence read data coverage).
[0127] In some embodiments, the classifier 130 may use a beta-binomial probability distribution to create the model 140b, although other probability distributions, such as a binomial probability distribution, a Gaussian distribution, and / or other probability modeling, may also be used.
[0128] For a beta-binomial probability distribution, the classifier 130 may use the dbetabinom function in the VGAM package of the R Project according to equation (6):
number
[0129] The classifier 130 may generate a second probability output L0 of the second model 140b according to equation (5):
number
number
[0130] It should be noted that the first model 140a and the second model 140b do not have to use the same probability distribution, as long as the first model 140a and the second model 140b each output a probability.
[0131] The classifier 130 can compare the first probability output of the first model 140a and the second probability output of the second model 140b to determine which probability output is more likely. For example, the classifier 130 can determine whether a somatic homozygous deletion or a somatic heterozygous deletion is more likely. In a particular example, the classifier 130 can use a log-likelihood ratio ("LLR") to create an LLR score based on the first probability output (probability of somatic homozygous deletion) and the second probability output (probability of somatic heterozygous deletion). In some embodiments, one of the first or second probability output can be used as a null probability, such that if the LLR score does not exceed a threshold cutoff score, the null probability is rejected. For example, the classifier 130 can compare the LLR score to a threshold cutoff score to determine whether the second probability output should be rejected. In other words, if the LLR score exceeds a threshold cutoff score, the classifier 130 can determine that the first probability output should be selected. In this example, the classifier 130 can make a prediction based on the comparison that a somatic homozygous deletion for the gene is present in the IUT111 sample.
[0132] In some examples, to reduce errors, model 140A or 140B can be used to determine the genotype of the sample for each SNP that overlaps with a given gene. If the germline SNP is not determined to be heterozygous, the given gene can be labeled "no call" and no somatic homozygous or heterozygous deletions are associated with the given gene.
[0133] Learning the Threshold Score Cutoff
[0134] In some embodiments, the threshold cutoff score can be customized for different genes or other genetic material being assayed. For example, the BRCA1 gene can be associated with a different threshold cutoff score than the BRCA2 gene. Other genes can be associated with custom threshold cutoff scores as well. In these embodiments, the classifier 130 can be trained to determine the threshold cutoff score. In some of these embodiments, the classifier 130 can be trained to determine the threshold cutoff score for a particular gene. For example, the classifier 130 can be trained using a simulation of somatic heterozygous deletion starting from a TND sample. For example, with reference to Figures 10A and 10B, a limit of blank (LoB), or maximum LLR score, is expected to be found when no homozygous deletion is present for a BRCA1 and BRCA2 negative sample. With reference to Figures 10A and 10B, 100,000 cases of somatic heterozygous deletion starting from a TND sample were simulated. The TF distribution observed in 28,000 samples was used as the TF to determine the LoB for BRCA1 and BRCA2. As illustrated, the threshold cutoff scores for comparison with the LLR score are 20.1 and 0 for BRCA1 and BRCA2, respectively. Thus, if a somatic deletion of BRCA1 is observed in a sample of IUT111 and the LLR score for BRCA1 in the sample of IUT111 is greater than 20.1, the classifier 130 can predict that the somatic deletion is a somatic homozygous deletion. Similarly, if a somatic deletion of BRCA1 is observed in a sample of IUT111 and the LLR score for BRCA2 in the sample of IUT111 is greater than 0, the classifier 130 can predict that the somatic deletion is a somatic homozygous deletion. It should be noted that other genes can be similarly simulated to determine the threshold cutoff scores.
[0135] In some embodiments, the model validator 132 can use simulated and / or clinical data to validate the results of the model 140. For example, the model validator 132 can consult the diagnostic outcome data store 150 and / or the clinical outcome data store 160 to validate predictions. The simulated results can model a panel of known samples to generate predictions of the state of the genetic material of those samples. These results can be used to validate the results of previous predictions and / or future predictions.
[0136] FIG. 3 illustrates a process 300 for predicting somatic homozygous or heterozygous deletions based on a trained model according to an embodiment of the present disclosure. Since there are various ways to perform the methods described herein, the process 300 is provided as an example. Although the process 300 is described as being primarily performed by the computer system 110 (via the processor 120) illustrated in FIG. 1, the process 300 may be performed or otherwise implemented by other systems or combinations of systems. Each block illustrated in FIG. 3 may further represent one or more processes, methods, or subroutines, one or more of the blocks may include machine-readable instructions stored on a non-transitory computer-readable medium and executed by a processor or other type of processing circuitry to perform one or more operations described herein. Various operations of the process 300 disclosed herein, or blocks executed by the systems disclosed herein, may be performed at the same or different times in the same or different geographic locations, e.g., countries, and / or by the same or different people.
[0137] At operation 302, the processor 120 can access germline SNP data from a set of samples including a TND sample. At operation 304, the processor 120 can determine a heterozygosity rate and a SD of the MAF based on the germline SNP data. At operation 306, the processor 120 can determine a first number of reads for each germline SNP site that support a determination that the sample from the individual contains a somatic homozygous deletion in the gene.
[0138] In operation 308, the processor 120 may generate a first output of a first model of the probability that the gene is associated with a somatic homozygous deletion based on the heterozygosity rate, the standard deviation (sd) of the MAF, the germline SNP data, and the number of reads. In operation 310, the processor 120 may determine the mean and standard deviation of the TFs based on the sample from the individual. In operation 312, the processor 120 may determine a second number of reads for each germline SNP site that supports a determination that the sample from the individual contains a somatic heterozygous deletion in the gene. In operation 314, the processor 120 may generate a second output of a second model of the probability that the gene is associated with a somatic heterozygous deletion based on the mean and SD of the TFs, the germline SNP data, and the second number of reads. In operation 316, the processor 120 may compare the first output and the second output. In operation 318, the processor 120 may determine whether the first output should be selected based on the comparison. In operation 320, the processor 120 may generate a probability that the gene contains a somatic homozygous deletion based on the determination of whether the first output should be selected or not.
[0139] The classifier 130 can apply various modeling techniques to predict the genetic state. The classifier 130 can use other modeling techniques as well. For example, Figures 9A and 9B illustrate a comparison of the results of different modeling techniques. Other probability techniques can be used as well.
[0140] Modeling other types of states of genetic material
[0141] The classifier 130 can model other types of states of genetic material. For example, the classifier 130 can predict various types of states of genetic material, such as CNV. Reference will now be made to FIG. 4 illustrating a process 400 for predicting a state of genetic material. Since there may be various ways to perform the methods described herein, the process 400 is provided as an example. Although the method 400 is described as being performed primarily by the computer system 110 (via the processor 120) illustrated in FIG. 1, the process 400 may be performed or otherwise performed by other systems or combinations of systems. Each block illustrated in FIG. 4 may further represent one or more processes, methods, or subroutines, one or more of the blocks may include machine-readable instructions stored on a non-transitory computer-readable medium and executed by a processor or other type of processing circuitry to perform one or more operations described herein. Various operations of the process 400 disclosed herein, or blocks performed by the systems disclosed herein, may be performed at the same or different times in the same or different geographic locations, e.g., countries, and / or by the same or different people.
[0142] The example described with respect to Figure 4 involves determining CNVs in a sample of IUT111. More specifically, the example can be used to determine copy number variance (such as amplification) in genetic material from a sample of IUT111. However, other types of states of genetic material can be determined in a similar manner, using alternative (competing) probabilities of different states and selecting the most likely probability.
[0143] At operation 402, the processor 120 may create a first model that models a first state of the genetic material. The first state may include a first CNV or other state. At operation 404, the processor 120 may create a second model that models a second state of the genetic material. The second state may include a second CNV or other state. At operation 406, the processor 120 may create a first score based on the first model. The first score may indicate a probability that the genetic material is in the first state.
[0144] At operation 408, the processor 120 may generate a second score based on the second model. The second score may indicate a probability that the genetic material is in the second state. At operation 410, the processor 120 may compare the first score and the second score. At operation 412, the processor 120 may generate a prediction that the genetic material is in the first state or the second state based on the comparison.
[0145] Similar to how the MAF for germline SNPs is used by classifier 130 to generate probabilities of somatic heterozygous and homozygous deletions, the MAF can be used to resolve the probability of CNV. For example, the MAF of a germline SNP in a sample where no CNV is detected can be used to determine whether the reads in the sample support a particular amplification.
[0146] FIG. 5 illustrates the types of somatic deletions according to embodiments of the present disclosure. Somatic homozygous deletions can occur in two ways: (1) germline cells have a single copy of the gene and somatic cells acquire a second deletion (LoD similar to detecting single copy amplifications). These can be detected based on coverage + no overlapping heterozygous SNPs. In some cases, germline cells cannot have a copy of the gene even though these were not observed. (2) The second way somatic homozygous deletions can occur is that germline cells have two copies of the gene and somatic cells lose both copies (this scenario was observed with a higher incidence). In some embodiments, in biallelic somatic copy number loss, the reference allele frequency of the germline heterozygous SNP is 0.5 in a mixture of germline and somatic cells. In the case of somatic LOH, the reference allele frequency is 0.5-0.5, depending on whether the reference allele is lost or retained in the cancer cells. * TF (tumor content) or 0.5+0.5 * TF.In some embodiments, in LOH, expected allele frequency can vary according to the proportion of tumor cells.Therefore, the system can distinguish between LOH and biallelic copy number loss based on calculated allele frequency compared with expected allele frequency of 0.5.
[0147] FIG. 6A illustrates exemplary plots 600(A)(1) and 600(A)(2) of BRCA1 homozygous deletions according to embodiments of the present disclosure. FIG. 6B illustrates exemplary plots 600(A)(1) and 600(A)(2) of BRCA2 heterozygous deletions according to embodiments of the present disclosure. With reference to plots 600(A)(1) and 600(B)(1), for a given cfDNA sample, normalized molecular coverage (y-axis) is represented across targeted probes (x-axis) sorted by genomic location. Chromosomal segregation is represented by vertical lines and identifiers provided at the bottom line of the plot. Regions without somatic copy number changes show molecular coverage approaching 2, while somatic deletions may be identified by molecular coverage levels below 2. With reference to plots 600(B)(1) and 600(B)(2), in the same sample, the MAF of known germline SNPs (y-axis) is represented relative to their genomic location (x-axis). Somatic deletions, as observed in the coverage plots in the top row, demonstrate a germline variant MAF close to 50% (see plot illustrated in FIG. 6A), while heterozygous deletions result in unbalanced germline variant MAF (see plot illustrated in FIG. 6B).
[0148] Figure 7A illustrates an exemplary plot of the incidence of heterozygous genotypes for known germline SNPs overlapping ATM, BRCA1 and BRCA2 genes observed in TND samples according to an embodiment of the present disclosure. Figure 7B illustrates an exemplary plot of MAF across TND samples according to an embodiment of the present disclosure.
[0149] FIG. 8A illustrates an exemplary plot of MAF values for BRCA1 according to an embodiment of the present disclosure. FIG. 8B illustrates an exemplary plot of MAF values for BRCA2 according to an embodiment of the present disclosure. FIGS. 8A and 8B show examples of MAFs (y-axis) of nine known germline SNVs for the three possible genotypes (homozygous alternative allele / heterozygous / homozygous reference allele) (x-axis) of each SNP. FIG. 9A illustrates an exemplary plot of a score comparison between a beta-binomial model and a binomial model for a BRCA2 panel according to an embodiment of the present disclosure. FIG. 9B illustrates an exemplary plot of a score comparison between a beta-binomial model and a Gaussian distribution model for a BRCA2 panel according to an embodiment of the present disclosure. FIG. 10A illustrates an exemplary plot of LLR score distribution for BRCA1 negative samples according to an embodiment of the present disclosure. FIG. 10B illustrates an exemplary plot of LLR score distribution for BRCA2 negative samples according to an embodiment of the present disclosure.
[0150] Figure 11A illustrates an exemplary plot of LoD of deletion for BRCA1 according to an embodiment of the present disclosure. Figure 11B shows an exemplary plot of LoD of loss of heterozygosity (LOH) for BRCA1 (interchangeably referred to herein as "heterozygous deletion") according to an embodiment of the present disclosure. Simulation: 100k cases of homozygous cell deletion starting from TND samples.
[0151] TFs used = TF distribution observed in 28,199 samples. LoD depends on two factors (two-step algorithm): (1) deletion detection sensitivity (based on coverage only): BRCA1 amplification / deletion mean cutoff = 0.05; and (2) ability to distinguish between homozygous and heterozygous somatic deletions (LLR test).
[0152] Figure 12A illustrates an exemplary plot of the LoD of deletions for BRCA2 according to an embodiment of the present disclosure. Figure 12B illustrates an exemplary plot of the LoD of LOH for BRCA2 according to an embodiment of the present disclosure. Simulation: 100k cases of homozygous cell deletions starting from TND samples.
[0153] TFs used = TF distribution observed in 28,199 samples.
[0154] The LoD depends on two factors (two-step algorithm): (1) deletion detection sensitivity (based on coverage alone): BRCA2 amplification / deletion mean cutoff = 0.09; and (2) the ability to distinguish between homozygous and heterozygous somatic deletions (LLR test).
[0155] FIG. 13 illustrates an exemplary plot of TF prevalence versus cancer type, according to an embodiment of the present disclosure.
[0156] Figure 14 illustrates an exemplary plot of density of LLR scores for BRCA1 and BRCA2 according to an embodiment of the present disclosure. A set of 28,000 training samples was randomly selected, with cutoffs of 2.5 and 0 (determined in LoB section), and samples were called with BRCA1 / 2 homozygous deletion. 387 and 994 samples showed somatic deletion for BRCA1 and BRCA2, respectively. Of these samples, 49 and 60 were called with homozygous deletion of BRCA1 and BRCA2, respectively.
[0157] FIG. 15 illustrates an exemplary chart of observed prevalence of BRCA2 homozygous deletions in a population of cancer types according to an embodiment of the present disclosure. FIG. 16 illustrates an exemplary chart of observed prevalence of BRCA1 homozygous deletions in a population of cancer types according to an embodiment of the present disclosure. FIG. 17 illustrates an example of BRCA2 homozygous deletions and potential clinical addressability according to an embodiment of the present disclosure. The plot illustrated in FIG. 17 is from “Integrative clinical genomics of advanced prostate cancer,” Cell 161:1215-1228 (2015) by Robinson D, Van Allen EM, Wu YM, Schultz N, Lonigro RJ, Mosquera JM, Montgomery B, Taplin ME, Pritchard CC, Attard G, et al. (“Robinson”), which is incorporated herein by reference in its entirety. Robinson shows that integrated analysis of both somatic and pathogenic germline alterations in BRCA2 identified 19 / 150 (12.7%) cases of BRCA2 loss, of which approximately 90% displayed biallelic loss. This was generally the result of somatic point mutations and losses of heterozygosity, as well as homozygous deletions. A clinical study evaluating poly(ADP-ribose) polymerase (PARP) inhibition in unselected individuals with mCRPC showed that multiple affected individuals in this study who experienced clinical benefit carried biallelic BRCA2 losses, providing further evidence of clinical addressability.
[0158] Figure 18A illustrates an exemplary plot of BRCA1 homozygous deletion according to an embodiment of the present disclosure. Figure 18B illustrates an exemplary plot of BRCA1 homozygous deletion according to an embodiment of the present disclosure. Figure 19A illustrates an exemplary plot of BRCA2 homozygous deletion according to an embodiment of the present disclosure. Figure 19B illustrates an exemplary plot of BRCA2 homozygous deletion according to an embodiment of the present disclosure. Figures 18A, 18B, 19A, and 19B are plots based on the human genome.
[0159] FIG. 20A illustrates an exemplary plot of BRCA1 biallelic somatic copy number loss according to an embodiment of the present disclosure. For purposes of this disclosure, the term "biallelic somatic copy number loss" will be used interchangeably with "homozygous deletion". FIG. 20B illustrates an exemplary plot of BRCA1 LOH according to an embodiment of the present disclosure. For purposes of this disclosure, the term "LOH" will be used interchangeably with "heterozygous deletion". FIG. 21A illustrates an exemplary plot of BRCA2 biallelic somatic copy number loss according to an embodiment of the present disclosure. FIG. 21B illustrates an exemplary plot of BRCA2 LOH according to an embodiment of the present disclosure. FIGs. 20A, 20B, 21A, and 21B are plots based on three (human) chromosomes. FIG. 22 illustrates a plot of the incidence of BRCA1 and BRCA2 somatic deletion according to an embodiment of the present disclosure.
[0160] Computer implementation
[0161] The method may be computer-implemented such that any or all of the steps described herein or in the appended claims, other than the wet chemistry steps, may be performed on a suitable programmed computer. The computer may be a mainframe, a personal computer, a tablet, a smartphone, a cloud, online data storage, remote data storage, etc. The computer may be operated in one or more locations.
[0162] Various operations of the method may utilize information and / or programs and result in results being stored on a computer-readable medium (e.g., a hard drive, secondary memory, external memory, a server; a database, a portable memory device (e.g., a CD-R, a DVD, a ZIP disk, a flash memory card), etc.).
[0163] The present disclosure also includes articles of manufacture for analyzing nucleic acid populations, including machine-readable media containing one or more programs that, when executed, implement the steps of the methods of the invention.
[0164] The present disclosure may be implemented in hardware and / or software. For example, different aspects of the present disclosure may be implemented in either client-side logic or server-side logic. The present disclosure or components thereof may be embodied in a fixed-medium program component that contains logic instructions and / or data that, when loaded into a suitably configured computing device, causes the device to perform according to the present disclosure. The fixed medium containing the logic instructions may be delivered to the viewer on the fixed medium for physical loading into the viewer computer, or the fixed medium containing the logic instructions may reside on a remote server that the viewer accesses via a communication medium to download the program component.
[0165] The present disclosure provides a computer control system programmed to implement the methods of the present disclosure. The processor 120 may include a single-core or multi-core processor, or multiple processors for parallel processing. The storage device 122 may include random access memory, read-only memory, flash memory, hard disk, and / or other types of storage. The computer system 110 may include a communication interface (e.g., a network adapter) for communicating with one or more other systems, as well as peripheral devices such as cache, other memory, data storage, and / or electronic display adapters. The components of the computer system 110 may communicate with each other via an internal communication bus, e.g., a motherboard. The storage device 122 may be a data storage unit (or data repository) for storing data. The computer system 110 may be operatively coupled to a computer network ("network") with the aid of a communication interface. The network may be the Internet, an Internet and / or an extranet, or an intranet and / or an extranet in communication with the Internet. The network is, in some cases, a telecommunications and / or data network. The network may include a local area network. The network may include one or more computer servers, which may enable distributed computing such as cloud computing. The network may implement a peer-to-peer network, which in some cases with the aid of computer system 110 may enable devices to be coupled to computer system 120 functioning as a client or a server.
[0166] The processor 120 can execute sequences of machine-readable instructions, which may be embodied in a program or software. The instructions may be stored in a memory location, such as the storage device 122. The instructions may be targeted to the processor 120, which may then be programmed or otherwise configured to implement the methods of the present disclosure. Examples of operations performed by the processor 120 may include fetch, decode, execute, and writeback.
[0167] The processor 120 may be part of a circuit, such as an integrated circuit. One or more other components of the system 100 may be included in the circuit. In some cases, the circuit may include an application specific integrated circuit (ASIC).
[0168] The storage device 122 may store files, such as drivers, libraries, and saved programs. The storage device 122 may store user data, such as user preferences and user programs. The computer system 110 may include one or more additional data storage units, in some cases located outside the computer system 110, for example, on a remote server that communicates with the computer system 110 via an intranet or the Internet.
[0169] The computer system 110 can communicate with one or more remote computer systems over a network. For example, the computer system 110 can communicate with a user's remote computer system. Examples of remote computer systems include a personal computer (e.g., a portable PC), a slate or tablet PC (e.g., Apple® iPad®, Samsung® Galaxy Tab), a telephone, a smartphone (e.g., Apple® iPhone®, Android® compatible, Blackberry®), or a personal digital assistant. A user can access the computer system 110 over a network.
[0170] The methods described herein may be implemented by machine (e.g., computer processor) executable code stored in an electronic storage location of computer system 110, such as, for example, storage device 122. The machine executable or machine readable code may be provided in the form of software. In use, the code may be executed by processor 905. In some cases, the code may be retrieved from storage unit 915 and stored in storage device 122 for easy access by processor 120.
[0171] The code may be precompiled and configured for use on a machine having a processor adapted to execute the code, or may be compiled during run-time. The code may be provided in a programming language that may be selected to render the code executable, in a precompiled or compiled fashion.
[0172] Aspects of the systems and methods provided herein, such as the computer system 110, may be embodied by programming. Various aspects of the technology may be thought of as a "product" or "article of manufacture," typically in the form of machine (or processor) executable code and / or associated data executed on or embodied in some type of machine-readable medium. The machine-executable code may be stored in an electronic storage unit, such as a memory (e.g., read-only memory, random access memory, flash memory) or a hard disk.
[0173] A "storage" type medium may include any or all of the tangible memory of a computer, processor, or the like, or associated modules thereof, such as various semiconductor memories, tape drives, disk drives, etc., that can provide non-transitory storage at any point in time for software programming. All or a portion of the software may be communicable over the Internet or various other telecommunications networks. Such communication may, for example, enable loading of the software from one computer or processor to another, such as from a management server or host computer to a computer platform of an application server. Thus, another type of medium that may bear software elements includes light waves, radio waves, and electromagnetic waves, such as those used across physical interfaces between local devices, through wired or optical landline networks, and through various air links. Physical elements that carry such waves, such as wired or wireless links, optical links, etc., may also be considered media bearing software. As used herein, non-transitory, tangible,
[0174] Unless limited to "storage" media, terms such as computer or machine "readable medium" refer to any medium that participates in providing instructions to a processor for execution.
[0175] Thus, a machine-readable medium such as a computer executable code can take many forms, including, but not limited to, a tangible storage medium, a carrier wave medium, or a physical transmission medium. Non-volatile storage media include optical or magnetic disks, such as any of the storage devices in any computer, such as may be used to implement the databases, etc., shown in the drawings. Volatile storage media include dynamic memory, such as the main memory of such a computer platform. Tangible transmission media include coaxial cables; copper wire and optical fiber, including the wires that comprise a bus within a computer system. Carrier wave transmission media can take the form of electric or electromagnetic signals, or acoustic or light waves, such as those generated during radio frequency (RF) infrared (IR) data communications. Thus, common forms of computer readable media include, for example, a floppy disk, a flexible disk, a hard disk, a magnetic tape, any other magnetic medium, a CD-ROM, a DVD or DVD-ROM, any other optical medium, punch cards, paper tape, any other physical storage medium having a pattern of holes, a RAM, a ROM, a PROM and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave transporting data or instructions, a cable or link transporting such a carrier wave, or any other medium from which a computer can read programming code and / or data. Many of these forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.
[0176] The computer system 110 may include or communicate with an electronic display 935 that includes a user interface (UI) for providing, for example, reports. Examples of UIs include, but are not limited to, graphical user interfaces (GUIs) and web-based user interfaces.
[0177] The methods and systems of the present disclosure may be implemented as one or more algorithms. The algorithms may be implemented as software when executed by the processor 120.
[0178] Sampling and Analysis Pipeline
[0179] The sample 101 may be any biological sample isolated from a subject. Samples may include body tissues, such as known or suspected solid tumors, whole blood, platelets, serum, plasma, stool, red blood cells, white blood cells or cells, endothelial cells, tissue biopsies, cerebrospinal fluid, synovial fluid, lymphatic fluid, ascites fluid, interstitial or extracellular fluid, gingival crevicular fluid, bone marrow, pleural fluid, cerebrospinal fluid, saliva, mucus, sputum, semen, sweat, and fluid in the spaces between cells, including urine. The samples are preferably body fluids, in particular blood and fractions thereof, and urine. Such samples may include nucleic acids. Such samples may also be referred to as nucleic acid samples. In some of these samples, the nucleic acids may be shed from tumors. The nucleic acids may include DNA and RNA, and may be in double-stranded and / or single-stranded form. In the example where the nucleic acid comprises RNA, the system and method described herein can determine somatic deletion in the gene of interest encoded by RNA by comparing the gene expression of the gene of interest relative to a reference gene (such as an endogenous control gene such as GAPDH) with a trained threshold calculated from a normal sample.The sample may be in the form originally isolated from the subject, or may be subjected to further processing to remove or add components such as cells, enrich one component relative to another, or convert one form of nucleic acid into another, for example, RNA into DNA, or single-stranded nucleic acid into double-stranded.Thus, for example, the body fluid for analysis is plasma or serum that contains cell-free nucleic acid, for example, cell-free DNA (cfDNA).
[0180] The volume of plasma may vary depending on the desired read depth for the sequenced region. Exemplary volumes are 0.4-40 ml, 5-20 ml, 10-20 ml. For example, the volume may be 0.5 ml, 1 ml, 5 ml, 10 ml, 20 ml, 30 ml, or 40 ml. The volume of sampled plasma may be 5 to 20 ml.
[0181] A sample can contain various amounts of nucleic acid containing genome equivalents. For example, a sample of about 30 ng of DNA can contain about 10,000 (10 4 ) haploid human genome equivalents, and in the case of cfDNA, approximately 200 billion (2 × 10 11 ) individual polynucleotide molecules. Similarly, a sample of about 100 ng of DNA may contain about 30,000 haploid human genome equivalents, or in the case of cfDNA, about 600 billion individual molecules.
[0182] The sample may include nucleic acids from different sources, for example, from cells and acellular sources. The sample may include nucleic acids with mutations. For example, the sample may include DNA with germline mutations and / or somatic mutations. The sample may include DNA with cancer-associated mutations (e.g., cancer-associated somatic mutations).
[0183] Exemplary amounts of cell-free nucleic acid in a sample prior to amplification range from about 1 fg to about 1 μg, for example, 1 pg to 200 ng, 1 ng to 100 ng, 10 ng to 1000 ng. For example, the amount can be up to about 600 ng, up to about 500 ng, up to about 400 ng, up to about 300 ng, up to about 200 ng, up to about 100 ng, up to about 50 ng, or up to about 20 ng of cell-free nucleic acid molecules. The amount can be at least 1 fg, at least 10 fg, at least 100 fg, at least 1 pg, at least 10 pg, at least 100 pg, at least 1 ng, at least 10 ng, at least 100 ng, at least 150 ng, or at least 200 ng of cell-free nucleic acid molecules. The amount can be up to 1 femtogram (fg), 10 fg, 100 fg, 1 picogram (pg), 10 pg, 100 pg, 1 ng, 10 ng, 100 ng, 150 ng, or 200 ng of cell-free nucleic acid molecules. The method can include obtaining from 1 femtogram (fg) to 200 ng.
[0184] A cell-free nucleic acid sample refers to a sample from a subject that contains cell-free nucleic acid. Cell-free nucleic acid is nucleic acid that is not contained within a cell or otherwise not bound to a cell. For example, a cell-free nucleic acid sample can include nucleic acid remaining in a sample after removing intact cells. Cell-free nucleic acid can refer to all unencapsulated nucleic acid that originates from a body fluid (e.g., blood, urine, CSF, etc.) from a subject. Cell-free nucleic acid includes DNA (cfDNA), RNA (cfRNA), and hybrids thereof, or fragments of any of these, including genomic DNA, mitochondrial DNA, circulating DNA, siRNA, miRNA, circulating RNA (cRNA), tRNA, rRNA, small nucleolar RNA (snoRNA), Piwi-interacting RNA (piRNA), long non-coding RNA (long ncRNA). Cell-free nucleic acid may be double-stranded, single-stranded, or hybrids thereof. Cell-free nucleic acid may be released into body fluids by secretion or cell death processes, such as necrosis and apoptosis of cells. Some cell-free nucleic acids are released into bodily fluids from cancer cells, e.g., circulating tumor DNA (ctDNA). Others are released from healthy cells. ctDNA can be fragmented DNA from unencapsulated tumors. Cell-free fetal DNA (cffDNA) is fetal DNA that circulates freely in the maternal bloodstream.
[0185] The cell-free nucleic acid or its associated proteins may have one or more epigenetic modifications, for example, the cell-free nucleic acid may be acetylated, 5-methylated, ubiquitinated, phosphorylated, sumoylated, ribosylated, and / or citrullinated.
[0186] The cell-free nucleic acids have an exemplary size distribution of about 100-500 nucleotides, with molecules of 110 to about 230 nucleotides representing about 90% of the molecules, with the mode in humans being about 168 nucleotides, and a second minor peak in the range of 240 to 440 nucleotides. The cell-free nucleic acids may be about 160 to about 180 nucleotides, or about 320 to about 360 nucleotides, or about 440 to about 480 nucleotides.
[0187] Cell-free nucleic acid can be isolated from body fluids by a partitioning step that separates the cell-free nucleic acid found in solution from intact cells and other insoluble components of the body fluid. Partitioning can include techniques such as centrifugation or filtration. Alternatively, cells in the body fluid can be lysed and the cell-free and cellular nucleic acids can be processed together. In general, after adding buffer and washing steps, the cell-free nucleic acid can be precipitated with alcohol. Additional clarification steps, such as silica-based columns, can be used to remove contaminants or salts. To optimize certain aspects of the procedure, such as yield, for example, non-specific bulk carrier nucleic acid can be added throughout the reaction.
[0188] After such processing, the sample may contain nucleic acids in various forms, including double-stranded DNA, single-stranded DNA, and single-stranded RNA. If necessary, single-stranded DNA and RNA may be converted to double-stranded form for inclusion in subsequent processing and analysis steps.
[0189] tag
[0190] In some embodiments, nucleic acid molecules (from a sample of polynucleotides) may be tagged with a sample index and / or a molecular barcode (commonly referred to as a "tag"). The tag may be incorporated into or otherwise joined to an adapter by chemical synthesis, ligation (e.g., blunt-end ligation or sticky-end ligation), or overlap extension polymerase chain reaction (PCR), among other methods. Such an adapter may ultimately be joined to a target nucleic acid molecule. In other embodiments, one or more amplification cycles (e.g., PCR amplification) are typically applied to introduce the sample index into the nucleic acid molecule using conventional nucleic acid amplification methods. Amplification may be performed in one or more reaction mixtures (e.g., multiple microwells in an array). The molecular barcode and / or sample index may be introduced simultaneously or in any sequential order. In some embodiments, the molecular barcode and / or sample index are introduced before and / or after the sequence capture step is performed. In some embodiments, only the molecular barcode is introduced before the probe capture, and the sample index is introduced after the sequence capture step is performed. In some embodiments, both the molecular barcode and the sample index are introduced before performing the probe-based capture step. In some embodiments, the sample index is introduced after the sequence capture step is performed. In some embodiments, the molecular barcode is incorporated into the nucleic acid molecule (e.g., cfDNA molecule) in the sample via an adaptor by ligation (e.g., blunt-end ligation or sticky-end ligation). In some embodiments, the sample index is incorporated into the nucleic acid molecule (e.g., cfDNA molecule) in the sample by overlap extension polymerase chain reaction (PCR). Typically, sequence capture protocols involve introducing a single-stranded nucleic acid molecule complementary to a targeted nucleic acid sequence, e.g., a coding sequence of a genomic region and a mutation in such region is associated with a type of cancer.
[0191] In some embodiments, the tag can be located at one or both ends of the sample nucleic acid molecule.In some embodiments, the tag is an oligonucleotide of predetermined or random or semi-random sequence.In some embodiments, the tag can be less than about 500, 200, 100, 50, 20, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 nucleotide length.The tag can be linked to the sample nucleic acid randomly or non-randomly.
[0192] In some embodiments, each sample is uniquely tagged with a sample index or a combination of sample indexes. In some embodiments, each nucleic acid molecule of a sample or sub-sample is uniquely tagged with a molecular barcode or a combination of molecular barcodes. In other embodiments, multiple molecular barcodes can be used such that the molecular barcodes are not necessarily unique to each other among the plurality (e.g., non-unique molecular barcodes). In these embodiments, the molecular barcodes are typically attached (e.g., by ligation) to individual molecules such that the combination of the molecular barcode and the sequence to which it may bind results in a unique sequence that can be tracked individually. Detection of the non-uniquely tagged molecular barcodes in combination with endogenous sequence information (e.g., beginning (start) and / or end (stop) portions corresponding to the sequence of the original nucleic acid molecule in the sample, subsequences of the sequence reads at one or both ends, length of the sequence reads, and / or length of the original nucleic acid molecule in the sample) typically allows for the assignment of a unique identity to a particular molecule. The length of the individual sequence reads, or the number of base pairs, is also used as needed to assign a unique identity to a given molecule. As described herein, fragments derived from a single strand of nucleic acid may be assigned a unique identity, thereby allowing for subsequent identification of fragments derived from the parent strand, and / or the complementary strand.
[0193] In some embodiments, molecular barcodes are introduced to molecules in a sample in a predicted ratio of a set of identifiers (e.g., a combination of unique or non-unique molecular barcodes). An exemplary format uses about 2 to about 1,000,000 different molecular barcodes, or about 5 to about 150 different molecular barcodes, or about 20 to about 50 different molecular barcodes ligated to both ends of a target molecule. Alternatively, about 25 to about 1,000,000 different molecular barcodes may be used. For example, 20-50 x 20-50 molecular barcodes may be used. Such a number of identifiers is typically sufficient to have a high probability (e.g., at least 94%, 99.5%, 99.99%, or 99.999%) that different molecules with the same start and stop points will receive different combinations of identifiers. In some embodiments, about 80%, about 90%, about 95%, or about 99% of the molecules have the same combination of molecular barcodes.
[0194] In some embodiments, the assignment of unique or non-unique molecular barcodes in the reactions is performed using methods and systems described, for example, in U.S. Patent Application Nos. 20010053519, 20030152490, and 20110160078, and U.S. Patent Nos. 6,582,908, 7,537,898, 9,598,731, and 9,902,992, each of which is incorporated herein by reference in its entirety. Alternatively, in some embodiments, different nucleic acid molecules of a sample may be identified using only intrinsic sequence information (e.g., start and / or stop positions, subsequences at one or both ends of the sequence, and / or length).
[0195] amplification
[0196] The sample nucleic acid flanked by adaptors can be amplified by PCR and other amplification methods that typically prime by binding primers to primer binding sites in adaptors flanking the DNA molecule to be amplified.The amplification method may involve cycles of extension, denaturation, and annealing as a result of thermal cycling, or may be isothermal, such as in transcription-mediated amplification.Other amplification methods include ligase chain reaction, strand displacement amplification, nucleic acid sequence-based amplification, and self-sustaining sequence-based replication.
[0197] Conventional nucleic acid amplification methods can be used to apply one or more rounds of amplification to introduce barcodes into nucleic acid molecules. Amplification may be performed in one or more reaction mixtures. Molecular tags and sample indexes / tags may be introduced simultaneously or in any sequential order. Molecular tags and sample indexes / tags may be introduced before and / or after sequence capture. In some cases, only molecular tags are introduced before probe capture, while sample indexes / tags are introduced after sequence capture. In some cases, both molecular tags and sample indexes / tags are introduced before probe capture. In some cases, sample indexes / tags are introduced after sequence capture. Typically, sequence capture involves introducing a single-stranded nucleic acid molecule complementary to a targeted sequence, e.g., a coding sequence of a genomic region, where mutations in such regions are associated with a type of cancer. Typically, amplification produces multiple non-uniquely or uniquely tagged nucleic acid amplicons with molecular tags and sample indexes / tags in sizes ranging from 200nt to 700nt, 250nt to 350nt, or 320nt to 550nt. In some embodiments, the amplicon has a size of about 300 nt. In some embodiments, the amplicon has a size of about 500 nt.
[0198] concentrated
[0199] In some embodiments, sequences are enriched prior to sequencing the nucleic acid. Enrichment is performed for a specific target region or non-specifically, as desired ("target sequence"). In some embodiments, targeted regions of interest can be enriched using a differential tiling and capture scheme with nucleic acid capture probes ("baits") selected for one or more bait set panels. Differential tiling and capture schemes generally use different relative concentrations of bait sets to differentially tile (e.g., at different "resolutions") across genomic regions associated with the baits, and are subjected to a constraint set (e.g., sequencer constraints, e.g., sequencing load, availability of each bait, etc.) to capture the targeted nucleic acid at a desired level for downstream sequencing. These targeted genomic regions of interest optionally include natural or synthetic nucleotide sequences of nucleic acid constructs. In some embodiments, biotin-labeled beads with probes for one or more regions of interest can be used to capture the target sequences, and then, as desired, these regions can be amplified for enrichment of the regions of interest.
[0200] Sequence capture typically involves the use of oligonucleotide probes that hybridize to target nucleic acid sequences. In certain embodiments, a probe set strategy involves tiling probes across the region of interest. Such probes can be, for example, about 60 to about 120 nucleotides long. Sets can have depths of about 2x, 3x, 4x, 5x, 6x, 8x, 9x, 10x, 15x, 20x, 50x, or more. In general, the effectiveness of sequence capture depends in part on the length of the sequence in the target molecule that is complementary (or nearly complementary) to the sequence of the probe. In some embodiments, the enriched population can be amplified prior to sequencing.
[0201] Sequencing Pipeline
[0202] With or without prior amplification, the adaptor-flanked sample nucleic acids can be subjected to sequencing, for example, by one or more sequencing devices 107. Sequencing methods include, for example, Sanger sequencing, high-throughput sequencing, pyrosequencing, sequencing by synthesis, single molecule sequencing, nanopore sequencing, semiconductor sequencing, sequencing by ligation, sequencing by hybridization, RNA-Seq (Illumina), Digital Gene Expression (Helicos), next-generation sequencing, Single Molecule Sequencing by Synthesis (SMSS) (Helicos), massively parallel sequencing, Clonal Single Molecule Array (Solexa), shotgun sequencing, Ion Torrent, Oxford Nanopore, Roche Genia, Maxim-Gilbert sequencing, primer walking, sequencing using PacBio, SOLiD, Ion Torrent, or Nanopore platforms. The sequencing reactions can be carried out in a variety of sample processing units, which can be multiple lanes, multiple channels, multiple wells, or other means of processing multiple sample sets substantially simultaneously. The sample processing units may also include multiple sample chambers capable of processing multiple runs simultaneously.
[0203] The sequencing reaction may be performed on one or more fragment types known to contain markers for cancer or other diseases. The sequencing reaction may be performed on any nucleic acid fragment present in the sample. The sequence reaction may provide at least 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, 99.9% or 100% sequence coverage of the genome. In other cases, the sequence coverage of the genome may be less than 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, 99.9% or 100%.
[0204] Simultaneous sequencing reaction can be carried out using multiplex sequencing.In some cases, cell-free polynucleotide can be sequenced at least 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000, 100,000 sequencing reaction.In other cases, cell-free polynucleotide can be sequenced less than 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000, 100,000 sequencing reaction.Sequencing reaction can be carried out sequentially or simultaneously.The following data analysis can be carried out for all or part of sequencing reaction. In some cases, data analysis may be performed on at least 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000, 100,000 sequencing reactions. In other cases, data analysis may be performed on less than 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 50000, 100,000 sequencing reactions. Exemplary read depths are 1000-80000 reads per locus (base).
[0205] Sequence analysis pipeline
[0206] In some embodiments, the nucleic acid in the sample can be contacted with a sufficient number of adapters containing molecular barcodes such that the probability that any two copies of the same nucleic acid molecule will receive the same combination of molecular barcodes from the adapters linked to both ends is low (e.g., less than 1 or 0.1%). The use of such adapters allows the identification of a family of nucleic acid sequences (sequence reads) resulting from a given nucleic acid molecule. For example, nucleic acid sequences that have the same start and stop points in a reference sequence and are linked to the same combination of molecular barcodes can be considered to be part of a family. In this way, a family represents the sequence of the amplification product of a given nucleic acid molecule in a sample, where the family members are the sequence reads obtained from the amplification products. The sequences of the family members can be compiled to derive a consensus nucleotide or a complete consensus sequence for the nucleic acid molecule in the original sample as modified by blunt end formation and adapter binding. In other words, it is determined that the nucleotide occupying a particular position of the nucleic acid in the sample is the consensus of the nucleotide occupying the corresponding position in the sequence of the family members. A family can include the sequence of one or both strands of double-stranded nucleic acid.When a family member includes the sequence of both strands from double-stranded nucleic acid, the sequence of one strand is converted to its complement in order to compile all sequences to derive consensus nucleotide or sequence.Some families only include the sequence of a single member.In this case, this sequence can be obtained as the sequence of the nucleic acid in the sample before amplification.Alternatively, families that only have the sequence of a single member can be excluded from the next analysis.
[0207] Nucleotide mutations in sequenced nucleic acids can be determined by comparing sequenced nucleic acids to a reference sequence. The reference sequence is often a known sequence, such as a known full genome sequence or partial genome sequence from a subject (e.g., a full genome sequence of a human subject). The reference sequence can be, for example, hG19 or hG38. The sequenced nucleic acid can represent a sequence determined directly for a nucleic acid in a sample, or a consensus of sequences of an amplification product of such a nucleic acid, as described above. Comparison can be performed at one or more designated positions of the reference sequence. A subset of sequenced nucleic acids can be identified that includes a position that corresponds to a designated position of the reference sequence when each sequence is maximally aligned. Within such a subset, the sequenced nucleic acid can be determined to include a nucleotide mutation, if any, at a designated position, and optionally, to include a reference nucleotide, if any, (i.e., the same as in the reference sequence). If the number of sequenced nucleic acids in the subset that includes a nucleotide variant exceeds a selected threshold, a variant nucleotide can be called at the designated position. The threshold may be a single number, e.g., at least 1, 2, 3, 4, 5, 6, 7, 9, or 10 sequenced nucleic acids in the subset containing the nucleotide variant, or the threshold may be a proportion, e.g., at least 0.5, 1, 2, 3, 4, 5, 10, 15, or 20, of the sequenced nucleic acids in the subset containing the nucleotide variant, among other probabilities. The comparison may be repeated for any designated position of interest in the reference sequence. Sometimes, the comparison may be performed for designated positions occupying at least about 20, 100, 200, or 300 consecutive positions in the reference sequence, e.g., about 20-500, or about 50-300 consecutive positions.
[0208] The methods can be used to identify the presence or absence of genetic events that can cause a condition, particularly cancer, in a subject, to characterize the condition (e.g., stage the cancer or determine the heterogeneity of the cancer), to monitor the response to treatment of the condition, and to provide a prognostic risk of development of the condition or the subsequent course of the condition.
[0209] The method of the present invention can be used to detect various cancers. Cancer cells, like most cells, can be characterized by the rate of turnover, where old cells die and are replaced by new cells. In general, dead cells in contact with vasculature in a given subject can release DNA or fragments of DNA into the bloodstream. This also applies to cancer cells at various stages of disease. Depending on the stage of the disease, cancer cells can also be characterized by various genetic abnormalities, such as copy number mutations and rare mutations. This phenomenon can be used to detect the presence or absence of cancer in an individual using the methods and systems described herein.
[0210] The types and number of cancers that may be detected may include blood cancer, brain cancer, lung cancer, skin cancer, nasal cancer, throat cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, skin cancer, intestinal cancer, rectal cancer, thyroid cancer, bladder cancer, kidney cancer, oral cancer, stomach cancer, solid tumors, heterogeneous tumors, homogeneous tumors, and the like.
[0211] Cancer can be detected from genetic alterations including mutations, rare mutations, indels, copy number variations, transversions, translocations, inversions, deletions, aneuploidy, partial aneuploidy, polyploidy, chromosomal instability, structural changes in chromosomes, gene fusions, chromosomal fusions, gene truncations, gene amplification, gene duplications, chromosomal damage, DNA damage, abnormal changes in nucleic acid chemical modifications, and abnormal changes in epigenetic patterns.
[0212] Genetic data can also be used to characterize specific forms of cancer. Cancers are often heterogeneous in both composition and staging. Genetic profile data can allow characterization of specific subtypes of cancer, which can be important in diagnosing or treating them. This information can also give the subject or practitioner clues regarding the prognosis of a particular type of cancer, allowing either the subject or practitioner to adopt treatment options consistent with disease progression. Some cancers progress and become more aggressive and genetically unstable. Other cancers can remain benign, inactive, or dormant. The systems and methods of the present disclosure can be useful in determining disease progression.
[0213] The analysis of the present invention is also useful in determining the effectiveness of a particular treatment option. If treatment is successful, more cancers may die and release DNA, so a successful treatment option may increase the amount of copy number variation or rare mutation detected in the subject's blood. In other cases, this may not occur. In another example, perhaps a particular treatment option may correlate with the genetic profile of cancer over time. This correlation may be useful in selecting a treatment. Furthermore, if cancer is observed to remit after treatment, the method of the present invention can be used to monitor the remaining disease or recurrence of disease.
[0214] The method of the present invention can also be used to detect genetic mutations in conditions other than cancer. Immune cells, such as B cells, can undergo rapid clonal expansion in the presence of certain diseases. Clonal expansion can be monitored using detection of copy number mutations, and certain immune conditions can be monitored. In this example, analysis of copy number mutations can be performed over time to generate a profile of how a certain disease may progress. Detection of copy number mutations or even rarer mutations can be used to determine how a pathogen population is changing during the course of infection. This can be particularly important during chronic infections, such as HIV / AIDs or hepatitis infections, where viruses can change life cycle states and / or mutate to more virulent forms during the course of infection. The method of the present invention can be used to determine or profile the host's body's rejection activity when immune cells attempt to destroy transplanted tissue, monitor the status of the transplanted tissue, and modify the course of treatment or prevention of rejection.
[0215] Furthermore, the method of the present disclosure can be used to characterize the heterogeneity of abnormal conditions in a subject, the method includes creating a genetic profile of extracellular polynucleotides in the subject, the genetic profile including a plurality of data obtained from the analysis of copy number variation and rare mutation. In some cases, including but not limited to cancer, the disease may be heterogeneous. Disease cells may not be identical. In the example of cancer, it is known that some tumors contain different types of tumor cells, some cells at different stages of cancer. In other examples, the heterogeneity may include multiple foci of disease. Furthermore, in the example of cancer, there may be multiple tumor foci, where one or more foci are probably the result of metastasis spreading from the primary site.
[0216] The method can be used to generate a profile, fingerprint or data set that is the sum of genetic information from different cells in a heterogeneous disease, which may include analysis of copy number variations and rare mutations, either alone or in combination.
[0217] The methods can be used to diagnose, predict, monitor or observe cancer or other diseases of fetal origin, i.e., these methodologies can be used in pregnant subjects to diagnose, predict, monitor or observe cancer or other diseases in prenatal subjects whose DNA and other polynucleotides may be circulating with maternal molecules.
[0218] Examples of highly accurate treatment
[0219] The improved diagnosis provided by the computer system 110 can result in accurate treatment plans that can be specified by the computer system 110 (and / or curated by a medical practitioner). For example, one type of accurate diagnosis and treatment can relate to genes in the homology directed repair (HRR) pathway.
[0220] Homologous recombination is a type of genetic recombination in which nucleotide sequences are exchanged between two similar or identical DNA molecules. It is most widely used by cells to precisely repair harmful breaks that occur in both strands of DNA, known as double-strand breaks (DSBs). HRR provides a mechanism for error-free removal of damage present in replicated (S and G2 phase) DNA to eliminate chromosomal breaks before cell division occurs. The leading model for how homologous recombination repairs double-strand breaks in DNA is the homologous recombination repair pathway, which mediates the double-strand break repair (DSBR) pathway and the synthesis-dependent strand annealing (SDSA) pathway. Germline and somatic defects in homologous recombination genes have been strongly associated with breast, ovarian and prostate cancer.
[0221] The number and type of variant nucleotides in a sample can provide an indication of the susceptibility of the subject providing the sample to treatment, i.e., therapeutic intervention. For example, various poly ADP-ribose polymerase (PARP) inhibitors have been shown to stop the growth of tumors from breast, ovarian and prostate cancer caused by genetic mutations in BRCA1 or BRCA2 genes. Some of these therapeutic agents can inhibit base excision repair (BER), which can compensate for the defect of HRR.
[0222] On the other hand, certain BRCA and HRR wild type patients cannot achieve clinical benefit from treatment with PARP inhibitors. Moreover, not all ovarian cancer patients with BRCA mutations respond to PARP inhibitors. Moreover, different types of mutations may indicate different treatments. For example, somatic heterozygous deletions in the HRR gene may indicate different treatments than somatic homozygous deletions. Thus, the state of genetic material may affect treatment. In one example, PARP inhibitors can be administered to individuals who carry somatic homozygous deletions in the HRR gene, but cannot be administered to individuals who carry wild type alleles or somatic heterozygous deletions in the HRR gene.
[0223] Nucleotide variations in sequenced nucleic acids can be determined by comparing the sequenced nucleic acids with a reference sequence. The reference sequence is often a known sequence, for example, a known full genome sequence or partial genome sequence from a subject, a full genome sequence of a human subject. The reference sequence may be hG19. The sequenced nucleic acid may represent a sequence determined directly for a nucleic acid in a sample, or a consensus of sequences of an amplification product of such a nucleic acid, as described above. Comparison may be performed at one or more designated positions of the reference sequence. A subset of sequenced nucleic acids may be identified that includes a position that corresponds to a designated position of the reference sequence when each sequence is maximally aligned. Within such a subset, the sequenced nucleic acid may be determined to include a nucleotide variation, if any, at a designated position, and optionally, to include a reference nucleotide, if any (i.e., the same as in the reference sequence). If the number of sequenced nucleic acids in the subset that includes a nucleotide variant exceeds a threshold value, a variant nucleotide may be called at the designated position. The threshold may be a single number, e.g., at least 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 sequenced nucleic acids in the subset containing the nucleotide variant, or the threshold may be a proportion, e.g., at least 0.5, 1, 2, 3, 4, 5, 10, 15, or 20, of the sequenced nucleic acids in the subset containing the nucleotide variant, among other probabilities. The comparison may be repeated for any designated position of interest in the reference sequence. Sometimes, the comparison may be performed for designated positions occupying at least 20, 100, 200, or 300 consecutive positions in the reference sequence, e.g., 20-500, or 50-300 consecutive positions.
[0224] EXAMPLES
[0225] The modeling described herein was applied to plasma samples from 28,199 patients with advanced solid tumors that were sequenced using a 73-gene next-generation sequencing ctDNA panel from Guardant Health, Inc.
[0226] Example results show a sensitivity of 95% in detecting deletions of the BRCA1 / 2 genes for samples exhibiting tumor content of 9%-11%. Detection limits for LOH and biallelic copy number loss are 11%-13%. The observed incidence of BRCA1 somatic deletions is greater than 3% in breast, colorectal, prostate and endometrial cancers. The observed incidence of BRCA2 somatic deletions is greater than 6% in breast, lung, prostate, head and neck cancer (HNSCC) and hepatocellular carcinoma.
[0227] In a cohort of 5,568 patients with cancers associated with classical HRD, somatic LOH and biallelic somatic copy number loss were detected in 2.7% of samples for BRCA1 and 8.0% of samples for BRCA2, aligning with previously reported tissue occurrences. BRCA1 and BRCA2 LOH were observed in 2.4% (134 / 5568) and 7.4% (415 / 5568) of classical homologous recombination deficient (HRD) cancers, including breast, ovarian, prostate, and pancreatic cancers. BRCA1 and BRCA2 biallelic somatic copy number loss was observed in 0.3% (19 / 5568) and 0.5% (31 / 5568) of this same group of HRD cancers. BRCA1 / 2 somatic LOH and biallelic somatic copy number loss can be accurately detected in ctDNA based on application of the model described herein. The ability to identify this therapeutically targetable genomic alteration by non-invasive ctDNA assessment has important clinical implications, particularly in patients whose disease requires tissue testing due to deep visceral locations, primarily metastases to bone and brain, such as in breast and prostate cancer.
[0228] All patent applications, websites, other publications, accession numbers, etc. cited above or below are incorporated by reference in their entirety for all purposes to the same extent as if each individual item was specifically and individually indicated to be incorporated by reference. Where different versions of a sequence are associated with an accession number at different times, it refers to the version associated with the accession number at the effective filing date of this application. By effective filing date, it refers to the earlier of the actual filing date for the accession number or the filing date of the priority application, if applicable. Similarly, where different versions of a publication, website, etc. are published at different times, it refers to the version published closest to the effective filing date of this application, unless otherwise specified. Any feature, step, element, embodiment, or aspect of this disclosure may be used in combination with any other, unless otherwise specified. Although the present disclosure has been described in some detail by way of illustration and example for clarity and understanding, it will be apparent that certain changes and modifications can be practiced that are within the scope of the appended claims. In certain embodiments, for example, the following items are provided: (Item 1) 1. A computer system for distinguishing between somatic homozygous and somatic heterozygous deletions of a gene in a sample not showing a germline deletion of said gene, comprising: generating a first model of allele counts representing the somatic homozygous deletion based on the location of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene according to a first probability distribution; generating a second model of the number of alleles in the sample that represents the somatic heterozygous deletion based on the location of the one or more germline SNPs with a second probability distribution; comparing a first output of the first model with a second output of the second model; Based on the comparison, a prediction is made that the somatic homozygous deletion for the gene is present in the sample. A processor programmed to A computer system comprising: (Item 2) 2. The computer system of claim 1, wherein the first model represents a first probability that the sample contains the somatic homozygous deletion and the second model represents a second probability that the sample contains the somatic heterozygous deletion. (Item 3) 2. The computer system of claim 1, wherein the first probability distribution is the same type of probability distribution as the second probability distribution. (Item 4) 2. The computer system of claim 1, wherein the processor is programmed to determine one or more parameters for input to the first probability distribution to create the first model. (Item 5) 5. The computer system of claim 4, wherein the first probability distribution comprises one of a beta-binomial distribution, a binomial distribution, or a normal distribution. (Item 6) To generate the first model of allele counts, the processor: determining the incidence of heterozygosity for said one or more germline SNPs in a training set of samples for input to said first probability distribution; 5. The computer system according to item 4, further programmed to: (Item 7) 7. The computer system of claim 6, wherein the training set of samples comprises a plurality of tumor not detected (TND) samples. (Item 8) To generate the first model of allele counts, the processor: determining the standard deviation of minor allele frequencies (MAFs) associated with the one or more germline SNPs in the training set of samples for input to the first probability distribution; 7. The computer system of item 6, further programmed to: (Item 9) To generate the first model, the processor: determining the number of molecules in said sample that support a mutant allele for input to said first probability distribution; 9. The computer system of item 8, further programmed to: (Item 10) To generate the first model, the processor: determining a total number of molecules in the sample for input to the first probability distribution; 10. The computer system of item 9, further programmed to: (Item 11) To generate the first model, the processor: Calculating a first likelihood of the allele count at the one or more germline SNP positions in the sample assuming the somatic homozygous deletion based on molecular coverage associated with the somatic homozygous deletion. 11. The computer system of item 10, further programmed to: (Item 12) To generate the second model, the processor: calculating a second likelihood of the allele count at the one or more germline SNP positions in the sample assuming the somatic heterozygous deletion based on molecular coverage associated with a somatic heterozygous deletion; 12. The computer system of item 11, further programmed to: (Item 13) To generate the second model, the processor: determining a mean estimated tumor content from the samples for input to the second probability distribution for the second model; 5. The computer system according to item 4, further programmed to: (Item 14) 14. The computer system of claim 13, wherein the tumor content is estimated based on sequence coverage information. (Item 15) To generate the second model, the processor: determining a standard deviation of the estimated tumor incidence from the sample for input to the second probability distribution for the second model; Item 14. The computer system of item 13, further programmed to: (Item 16) The processor, Access multiple samples, identifying a set of samples from said plurality of samples that contain germline deletions; filtering and outputting the set of samples from the plurality of samples; Identifying the presence of the somatic homozygous deletion or the somatic heterozygous deletion from among the filtered plurality of samples. 2. The computer system of claim 1, further programmed to: (Item 17) 2. The computer system of claim 1, wherein the first output comprises a first probability for the presence of the somatic homozygous deletion and the second output comprises a second probability for the presence of the somatic heterozygous deletion. (Item 18) To compare the first output of the first model with the second output of the second model, the processor: Executing a log-likelihood function based on the first output and the second output. 15. The computer system of item 14, further programmed to: (Item 19) 2. The computer system of item 1, wherein the genes include one of BRCA1, BRCA2, and ATM. (Item 20) generating a first probability that a gene in the sample contains a somatic homozygous deletion; generating a second probability that the gene in the sample contains a somatic heterozygous deletion; comparing the first probability to the second probability; making a prediction as to whether said sample contains said somatic homozygous deletion or said somatic heterozygous deletion; A processor programmed to A system including: (Item 21) generating a first probability that the genetic material in the sample comprises a first state; generating a second probability that the genetic material in the sample comprises a second state; comparing the first probability to the second probability; making a prediction as to whether said sample contains said first state or said second state; A processor programmed to A system including: (Item 22) 22. The system of claim 21, wherein the first state comprises a somatic homozygous deletion determined based on a first model, and the second state comprises a somatic heterozygous deletion determined based on a second model. (Item 23) 22. The system of claim 21, wherein the first state comprises a first copy number variant (CNV) and the second state comprises a second CNV that is different from the first CNV. (Item 24) 24. The system of claim 23, wherein the first CNV and / or the second CNV is associated with a deleterious condition. (Item 25) To generate the first probability, the processor: accessing the location of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene; determining the standard deviation of the minor allele frequency (MAF) associated with said one or more germline SNPs in a training set of samples; 24. The system of claim 23, programmed to: (Item 26) To generate the first probability, the processor: determining the standard deviation of minor allele frequencies (MAFs) associated with said one or more germline SNPs in said training set of samples; 24. The system of claim 23, programmed to: (Item 27) 1. A processor-implemented method comprising: generating, by the processor, a first model of allele counts representing somatic homozygous deletions based on the positions of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene according to a first probability distribution; generating, by the processor, a second model of allele counts in the sample representing somatic heterozygous deletions based on the positions of the one or more germline SNPs with a second probability distribution; comparing, by the processor, a first output of the first model and a second output of the second model; generating, by said processor, a prediction that said somatic homozygous deletion for said gene is present in said sample based on said comparison; A method comprising: (Item 28) 1. A processor-implemented method comprising: generating, by said processor, a first probability that a gene in the sample contains a somatic homozygous deletion; generating, by the processor, a second probability that the gene in the sample contains a somatic heterozygous deletion; comparing, by the processor, the first probability to the second probability; generating, by said processor, a prediction of whether said sample contains said somatic homozygous deletion or said somatic heterozygous deletion based on said comparison; A method comprising: (Item 29) 1. A processor-implemented method comprising: generating, by the processor, a first probability that genetic material in the sample comprises a first state; generating, by the processor, a second probability that the genetic material in the sample comprises a second state; comparing, by the processor, the first probability to the second probability; making a prediction of whether the sample contains the first state or the second state based on the comparison; A method comprising: (Item 30) A method for administering to a subject determined to have a somatic homozygous deletion based on any preceding item, a therapeutic intervention effective to treat a cancer associated with said somatic homozygous deletion. (Item 31) 31. The method of claim 30, wherein the therapeutic intervention comprises a poly ADP ribose polymerase (PARP) inhibitor. (Item 32) 31. The method of claim 30, wherein the therapeutic intervention comprises a base excision repair (BER) inhibitor. (Item 33) A method for administering to a subject determined to have genetic material in a particular state based on any preceding item, a therapeutic intervention effective to treat a disease associated with said genetic material state. (Item 34) A method for administering a therapeutic intervention to exclude poly ADP ribose polymerase (PARP) inhibitors to a subject determined to not have a somatic homozygous deletion based on any preceding item. (Item 35) 30. The method or system of any one of the preceding items, further comprising the step of generating a report optionally comprising information regarding the status of the gene and / or genetic material in the sample and / or information derived therefrom. (Item 36) 36. The method or system of claim 35, further comprising a step of communicating the report to the subject from whom the sample originated or to a third party, such as a medical professional. (Item 37) 1. A method for determining whether a somatic deletion of a gene in a sample is a somatic homozygous deletion or a somatic heterozygous deletion, said sample not exhibiting a germline deletion of said gene, said method comprising: a) determining a reference allele frequency of a germline single nucleotide polymorphism (SNP) in said gene in said sample; b) said somatic deletion (i) a somatic homozygous deletion if the reference allele frequency of the germline SNP in the sample is within a predetermined range observed in a tumor-free detection (TND) sample, depending on the genotype of the SNP; or (ii) a somatic heterozygous deletion if the reference allele frequency of the germline SNP in the sample deviates from the predetermined range by a specified amount. and classifying the A method comprising: (Item 38) Item 38. The method according to item 37, wherein the predetermined range is 0 to 0.01, 0.45 to 0.55, or 0.99 to 1. (Item 39) determining a putative tumor content in the sample; calculating the defined amount based on the estimated tumor content; 38. The method of claim 37, further comprising:
Claims
1. 1. A computer system for distinguishing between somatic homozygous and somatic heterozygous deletions of a gene in a sample not showing a germline deletion of said gene, comprising: creating a first model of allele counts representing a first probability that the sample contains the somatic homozygous deletion based on a heterozygous occurrence rate of one or more germline single nucleotide polymorphisms (SNPs) associated with the gene in a training set of samples according to a first probability distribution and based on a standard deviation of minor allele frequencies (MAFs) associated with the one or more germline SNPs in the training set of samples, wherein to create the first model, a number of molecules in the sample supporting the presence of the somatic homozygous deletion and a total number of molecules in the sample are determined, and wherein the training set of samples includes a plurality of tumor not detected (TND) samples; creating a second model of allele counts in the sample that represents a second probability that the sample contains the somatic heterozygous deletion based on the occurrence of heterozygosity at the one or more germline SNPs in the training set of samples and based on the standard deviation of minor allele frequencies (MAFs) associated with the one or more germline SNPs in the training set of samples according to a second probability distribution, wherein a mean of tumor incidence estimated from the samples and a standard deviation of tumor incidence estimated from the samples are determined to create the second model; comparing a first output of the first model with a second output of the second model, wherein the first output comprises the first probability for the presence of the somatic homozygous deletion and the second output comprises the second probability for the presence of the somatic heterozygous deletion; making a prediction that the somatic homozygous deletion for the gene is present in the sample based on the comparison between the first probability and the second probability, where if the first probability is greater than the second probability, then the prediction that the somatic homozygous deletion for the gene is present in the sample. A processor programmed to A computer system comprising:
2. (a) the first probability distribution is the same type of probability distribution as the second probability distribution; (b) to create the first model, the processor is programmed to determine one or more parameters for input to the first probability distribution; and / or (c) the first probability distribution comprises a probability distribution comprising one of a beta-binomial distribution, a binomial distribution, or a normal distribution; The computer system of claim 1.
3. To generate the first model, the processor: Calculating a first likelihood of the allele count at the one or more germline SNP positions in the sample assuming the somatic homozygous deletion based on molecular coverage associated with a somatic homozygous deletion.
3. The computer system of claim 1, further programmed to:
4. To generate the second model, the processor: Calculating a second likelihood of the allele count at the one or more germline SNP positions in the sample assuming the somatic heterozygous deletion based on molecular coverage associated with a somatic heterozygous deletion. The computer system of any one of claims 1 to 3, further programmed to:
5. The computer system of claim 1 , wherein the tumor content is estimated based on sequence coverage information.
6. The processor, Access multiple samples, identifying a set of samples from said plurality of samples that contain germline deletions; filtering and outputting the set of samples from the plurality of samples; Identifying the presence of the somatic homozygous deletion or the somatic heterozygous deletion from among the filtered plurality of samples. The computer system of any one of claims 1 to 5, further programmed to:
7. The processor, performing a log-likelihood function based on the first output and the second output; The computer system of any one of claims 1 to 6, further programmed to:
8. The computer system of any one of claims 1 to 7, wherein the genes include one of BRCA1, BRCA2, and ATM.
Citation Information
Patent Citations
Analysis of genetic variants
US20180218113A1
Detecting cross-contamination in sequencing data
US20180373832A1
Methods of detecting somatic and germline variants in impure tumors
WO2018144782A1
