Models for Targeted Sequencing
A processing system uses a Bayesian inference based model for targeted sequencing or variant calling. In an embodiment, the processing system generates candidate variants of a cell free nucleic acid sample. The processing system determines likelihoods of true alternate frequencies for each of the candidate variants in the cell free nucleic acid sample and in a corresponding genomic nucleic acid sample. The processing system filters or scores the candidate variants by the model using at least the likelihoods of true alternate frequencies. The processing system outputs the filtered candidate variants, which may be used to generate features for a predictive cancer or disease model.
1 . (canceled)
2 . A method comprising:
accessing sequence reads of a cell free nucleic acid sample of a subject obtained from a first source of the subject, the sequence reads of the cell free nucleic acid sample comprising first depths and first alternate depths for a plurality of positions on a reference allele;
accessing sequence reads of a genomic nucleic acid sample of a subject obtained from a second source of the subject different than the first source of the subject, the sequence reads of the genomic nucleic acid sample comprising second depths and second alternate depths for the plurality of positions on a reference allele;
determining a first likelihood of true alternate frequency of the cell free nucleic acid sample by applying a first function to model the first alternate depths for the cell free nucleic acid sample and adding a first noise level to an output for the first function, wherein the first noise level describes expected noise rates of mutations in cell free nucleic acid samples per position of the plurality of positions on the reference allele;
determining a second likelihood of true alternate frequency of the genomic nucleic acid sample by applying a second function to model the second alternate depths of the genomic nucleic acid sample and adding a second noise level to an output for the second function, wherein the second noise level describes expected noise rates of mutations in genomic nucleic acid samples per position of the plurality of positions on the reference allele; and
outputting one or more candidate variants of the cell free nucleic acid sample based on the first likelihood of the true alternate frequency of the cell free nucleic acid sample and the second likelihood of the true alternate frequency of the genomic nucleic acid sample.
3 . The method of claim 2 , wherein the first function is a first Poisson distribution function parameterized by a first product of the first depths and a true alternate frequency of the cell free nucleic acid sample, and wherein the second function is a second Poisson distribution function parameterized by a second product of the second depths and a true alternate frequency of the genomic nucleic acid sample.
4 . The method of claim 2 , wherein outputting the one or more candidate variants of the cell free nucleic acid sample further comprises:
determining a probability that the true alternate frequency of the cell free nucleic acid sample is greater than a function of the true alternate frequency of the genomic nucleic acid sample, wherein the probability represents a confidence level that mutations from the first sequence reads from the cell free nucleic acid sample are not found in the second sequence reads from the genomic nucleic acid sample; and
filtering the plurality of candidate variants based on the determined probability; and
outputting the filtered candidate variants.
5 . The method of claim 4 , wherein determining the probability comprises:
determining a joint likelihood of the first likelihood and the second likelihood by:
determining a cumulative sum of one of the first and second likelihoods; and
determining an integral of the other of the first and second likelihoods.
6 . The method of claim 2 , further comprising:
determining the first noise level of mutations with respect to the cell free nucleic acid samples using a third function parameterized by first parameters; and
determining the second noise level of mutations with respect to the genomic nucleic acid samples using a fourth function parameterized by second parameters.
7 . The method of claim 6 , wherein the first and second parameters represent parameters of distributions that encode noise levels of mutations with respect to a given position of a sequence read.
8 . The method of claim 2 , further comprising:
collecting the cell free nucleic acid sample from a blood sample of the subject; and
performing enrichment on the cell free nucleic acid sample to generate the first sequence reads.
9 . A non-transitory computer readable storage medium storing computer program instructions that when executed by a processor cause the processor to:
access sequence reads of a cell free nucleic acid sample of a subject obtained from a first source of the subject, the sequence reads of the cell free nucleic acid sample comprising first depths and first alternate depths for a plurality of positions on a reference allele;
access sequence reads of a genomic nucleic acid sample of a subject obtained from a second source of the subject different than the first source of the subject, the sequence reads of the genomic nucleic acid sample comprising second depths and second alternate depths for the plurality of positions on a reference allele;
determine a first likelihood of true alternate frequency of the cell free nucleic acid sample by applying a first function to model the first alternate depths for the cell free nucleic acid sample and adding a first noise level to an output for the first function, wherein the first noise level describes expected noise rates of mutations in cell free nucleic acid samples per position of the plurality of positions on the reference allele;
determine a second likelihood of true alternate frequency of the genomic nucleic acid sample by applying a second function to model the second alternate depths of the genomic nucleic acid sample and adding a second noise level to an output for the second function, wherein the second noise level describes expected noise rates of mutations in genomic nucleic acid samples per position of the plurality of positions on the reference allele; and
output one or more candidate variants of the cell free nucleic acid sample based on the first likelihood of the true alternate frequency of the cell free nucleic acid sample and the second likelihood of the true alternate frequency of the genomic nucleic acid sample.
10 . The non-transitory computer readable storage medium of claim 9 , wherein the first function is a first Poisson distribution function parameterized by a first product of the first depths and a true alternate frequency of the cell free nucleic acid sample, and wherein the second function is a second Poisson distribution function parameterized by a second product of the second depths and a true alternate frequency of the genomic nucleic acid sample.
11 . The non-transitory computer readable storage medium of claim 9 , wherein instructions for outputting the one or more candidate variants of the cell free nucleic acid sample further cause the processor to:
determine a probability that the true alternate frequency of the cell free nucleic acid sample is greater than a function of the true alternate frequency of the genomic nucleic acid sample, wherein the probability represents a confidence level that mutations from the first sequence reads from the cell free nucleic acid sample are not found in the second sequence reads from the genomic nucleic acid sample; and
filter the plurality of candidate variants based on the determined probability; and
output the filtered candidate variants.
12 . The non-transitory computer readable storage medium of claim 11 , wherein instructions for determining the probability further cause the processor to:
determine a joint likelihood of the first likelihood and the second likelihood by:
determining a cumulative sum of one of the first and second likelihoods; and
determining an integral of the other of the first and second likelihoods.
13 . The non-transitory computer readable storage medium of claim 9 , further comprising instructions that cause the processor to:
determine the first noise level of mutations with respect to the cell free nucleic acid samples using a third function parameterized by first parameters; and
determine the second noise level of mutations with respect to the genomic nucleic acid samples using a fourth function parameterized by second parameters.
14 . The non-transitory computer readable storage medium of claim 13 , wherein the first and second parameters represent parameters of distributions that encode noise levels of mutations with respect to a given position of a sequence read.
15 . The non-transitory computer readable storage medium of claim 9 , further comprising instructions that cause the processor to:
collect the cell free nucleic acid sample from a blood sample of the subject; and
perform enrichment on the cell free nucleic acid sample to generate the first sequence reads.
16 . A system comprising a processor and a computer memory, the computer memory storing computer program instructions that when executed by a processor cause the processor to:
access sequence reads of a cell free nucleic acid sample of a subject obtained from a first source of the subject, the sequence reads of the cell free nucleic acid sample comprising first depths and first alternate depths for a plurality of positions on a reference allele;
access sequence reads of a genomic nucleic acid sample of a subject obtained from a second source of the subject different than the first source of the subject, the sequence reads of the genomic nucleic acid sample comprising second depths and second alternate depths for the plurality of positions on a reference allele;
determine a first likelihood of true alternate frequency of the cell free nucleic acid sample by applying a first function to model the first alternate depths for the cell free nucleic acid sample and adding a first noise level to an output for the first function, wherein the first noise level describes expected noise rates of mutations in cell free nucleic acid samples per position of the plurality of positions on the reference allele;
determine a second likelihood of true alternate frequency of the genomic nucleic acid sample by applying a second function to model the second alternate depths of the genomic nucleic acid sample and adding a second noise level to an output for the second function, wherein the second noise level describes expected noise rates of mutations in genomic nucleic acid samples per position of the plurality of positions on the reference allele; and
output one or more candidate variants of the cell free nucleic acid sample based on the first likelihood of the true alternate frequency of the cell free nucleic acid sample and the second likelihood of the true alternate frequency of the genomic nucleic acid sample.
17 . The system of claim 16 , wherein the first function is a first Poisson distribution function parameterized by a first product of the first depths and a true alternate frequency of the cell free nucleic acid sample, and wherein the second function is a second Poisson distribution function parameterized by a second product of the second depths and a true alternate frequency of the genomic nucleic acid sample.
18 . The system of claim 16 , wherein instructions for outputting the one or more candidate variants of the cell free nucleic acid sample further cause the processor to:
determine a probability that the true alternate frequency of the cell free nucleic acid sample is greater than a function of the true alternate frequency of the genomic nucleic acid sample, wherein the probability represents a confidence level that mutations from the first sequence reads from the cell free nucleic acid sample are not found in the second sequence reads from the genomic nucleic acid sample; and
filter the plurality of candidate variants based on the determined probability; and
output the filtered candidate variants.
19 . The system of claim 18 , wherein instructions for determining the probability further cause the processor to:
determine a joint likelihood of the first likelihood and the second likelihood by:
determining a cumulative sum of one of the first and second likelihoods; and
determining an integral of the other of the first and second likelihoods.
20 . The system of claim 16 , further comprising instructions that cause the processor to:
determine the first noise level of mutations with respect to the cell free nucleic acid samples using a third function parameterized by first parameters; and
determine the second noise level of mutations with respect to the genomic nucleic acid samples using a fourth function parameterized by second parameters.
21 . The system of claim 16 , further comprising instructions that cause the processor to:
collect the cell free nucleic acid sample from a blood sample of the subject; and
perform enrichment on the cell free nucleic acid sample to generate the first sequence reads.