Manchester, UK
Last Active
Manchester, UK


  • @mouradov Thanks for your reply. I too suspect the same. Trying to build quality distrib. plot from mpileup. Would post again if I am able to pin-point something relevant.
  • Ah! Very useful link @Geraldine_VdAuwera . In fact I did ran mutect by removing the parameter for 'normal' and received OK results. Just that the VCF's last col. (designated for the 'normal') had '0' for all tag:pair values. I thought this might not…
  • @Geraldine_VdAuwera Many thanks for the reply. Given VarScan's discrepancy I would have liked to use MuTect but am limited by lack of paired germline for most of my amplicon seq. samples. I would test HaplotypeCaller for these and Freebayes too. Tha…
  • hi @mouradov, Were you able to sort your issue. I am also analyzing amplicon seq. data and have run both VarScan and MuTect. Both are reporting read-depth that is wildly different from what I see in IGV; i.e. much lesser read-depth. ** I have NOT …
  • Oh, thats very interesting! Many thanks. I will try this. Will surely streamline my steps.
  • @Geraldine_VdAuwera Thank you for the suggestion. I will try it. Though it would take a lot of hand-holding for the data. But this should be able to pull out the ones affected by DP. I have used this approach in targeted MiSeq data for low cov. reg…
  • @Sheila Hi, I appended read depths to the variants called by HC and the min. for the variant allele is 1 (filtered depth of course). So I am little relieved that HC wouldn't do a blind refusal to call at low depth. What started my inquiry was failu…
  • hi GATK dev., I am using the GATK best practices for var-calling from RNA-seq and I was wondering if I could specify the minimum read depth in Haplotype Caller. I went through the documentation but couldn't find anything similar. RNA-seq data, by d…
  • hi Geraldine, Is the documentation for MuTect coming out soon? I am processing a lot of tumor normal pairs and to rerun them in future (when I get to know the tune-able parameters from the documentation) would be troublesome. A list of the parameter…