A lot of calls when doing calling on a "dirty" sample

[Milan Domazet]

Hello everyone,

First of all I want to say that I am very happy about Strelka's performance. It works fast and it provides good calls, but I had only one issue I could notice. I did some tests with a couple of somatic variant callers in order to make a good somatic pipeline. I noticed some issues with a "dirty" sample (a lot of low coverage SNPs all around). Strelka called a lot of those SNPs (around 300k) while other callers (Varscan and VarDict) were not even close (both around 1500). Is there anything that could be done about this, setting some filters for example? I find it strange that Strelka was the only caller going so high with the mutation number, so I was just wondering if that was something already known?

Thank you in advance!

Cheers,
Milan 

[B Bai]

Hi Milan,

I have same issue. Did you fix it?  Could the author comment on this? It seems that Strelka can detect much more mutations when AF is low (for example AF below 0.1). How can we filter it? Thank you very much!

Baoyan

[Milan Domazet]

Hello Baoyan,

sadly I did not find any issue for this. It would be good if we could get a hint from the tools author.

Thanks in advance,
Milan