Remove batch effect from exome data

Question

[Alternative]

Dear all,

We run targeted exome sequencing for ~100 samples divided in 6 sequencing batches. We noticed a batch effect whereby the samples coming from 2 of the 6 batches are heavily mutated. All samples were sequenced using the same protocol.

Any idea on how to remove batch effect from exome sequencing (targetted) samples.

Thanks for your help

[Pierre Lindenbaum]

did you look at any BAM QC ?

[Alternative]

Yes, I just run some:

Duplication is very high but expected since we have targeted sequencing.

Additionally, since samples correspond to PPFE, I checked in my VCFs the count of C > T conversions expected to be enriched for such samples. There, I found that the 2 over-mutated batches are the one highly enriched for C > T. This sounds like the reason now.

I believe the question now is: How can we remove/correct batch effects related to C > T conversions / PPFE from VCF files?

Thanks,