We, like many groups doing NGS, particularly on human samples, safeguards against sample mixups. However, we are constantly looking for new approaches and tools to improve our ability to:
- Catch sample identity problems
- Find low-level contamination in one sample by others
I am interested in hearing what tools or approaches folks are trying to accomplish the tasks above, especially those relying on NGS data alone (not SNP arrays, etc.).
•
written
by
Sean Davis
1 answer
VerifyBamID springs to mind: http://genome.sph.umich.edu/wiki/VerifyBamID
"verifyBamID is a software that verifies whether the reads in particular file match previously known genotypes for an individual (or group of individuals), and checks whether the reads are contaminated as a mixture of two samples. verifyBamID can detect sample contamination and swaps when external genotypes are available. When external genotypes are not available, verifyBamID still robustly detects sample swaps."
We up-front genotype all samples on a SNP platform and compare that at the other end from genotypes from e.g. exome sequencing. Obviously we have additional checks for inheritance sanity for family studies. We sometimes will receive familial trios with the wrong labels on, which are nice to be able to deconvolute simply, but also show that the mix up didn't happen in the lab.
Paired tumour/normals samples are obviously a deal breaker for this approach, so there is physical separation of tumour/normals in the lab so that they are never processed together in batches of library preps.
Log in to answer this question.
Hi Sean! Did your group ever explore options for this sort of QC? We are currently using VerifyBamID for both germline and tumor/normal data, but we are concerned with the possibility that it will erroneously interpret somatic copy number changes in tumor samples as contamination. Thinking about comparing to Conpair or GATK's CalculateContamination.