Detecting cross-contamination in sequencing data

A non-transitory computer-readable storage medium storing computer program instructions executable by a processor of a system to perform steps for identifying contamination in a test sample, the instructions, when executed by the processor, causing the processor to:

• receive a plurality of sequence read pairs comprising a plurality of forward strand sequence reads and a plurality of reverse strand sequence reads, wherein:
• each of the plurality of sequence read pairs are obtained from methylation-aware sequencing, and
• each of the plurality of sequence read pairs comprise at least one single nucleotide polymorphism (SNP);
• filter the plurality of sequence read pairs to generate a population of sequence read pairs by:
• filtering forward strand sequence reads according to a first ruleset describing forward strand sequence reads of the test sample that indicate contamination, and
• filtering reverse strand sequence reads according to a second ruleset describing reverse strand sequence reads of the test sample that indicate contamination;
• determine a prior contamination probability for each SNP of the population of sequence read pairs based on a minor allele frequency for each SNP;
• apply a contamination model including at least one likelihood test to a sequence read pair of the population using the contamination probabilities for the SNPs in that sequence read pair, each likelihood test configured to produce a test contamination probability representing the likelihood that the sequence read pair indicates a contamination in the test sample;
• and
• identify the contamination in the test sample when the test contamination probability is above a likelihood threshold.

A system for identifying contamination in a test sample, the system comprising:

• one or more memories storing computer program instructions for identifying contamination in the test sample;
• one or more processors configured to execute the computer program instructions for identifying contamination, the computer program instructions, when executed, causing the processor to:
• receive a plurality of sequence read pairs comprising a plurality of forward strand sequence reads and a plurality of reverse strand sequence reads, wherein:
• each of the plurality of sequence read pairs are obtained from methylation-aware sequencing, and
• each of the plurality of sequence read pairs comprise at least one single nucleotide polymorphism (SNP);
• filter the plurality of sequence read pairs to generate a population of sequence read pairs by:
• filtering forward strand sequence reads according to a first ruleset describing forward strand sequence reads of the test sample that indicate contamination, and
• filtering reverse strand sequence reads according to a second ruleset describing reverse strand sequence reads of the test sample that indicate contamination;
• determine a prior contamination probability for each SNP of the population of sequence read pairs based on a minor allele frequency for each SNP;
• apply a contamination model including at least one likelihood test to a sequence read pair of the population using the contamination probabilities for the SNPs in that sequence read pair, each likelihood test configured to produce a test contamination probability representing the likelihood that the sequence read pair indicates a contamination in the test sample;
• and
• identify the contamination in the test sample when the test contamination probability is above a likelihood threshold.