Entering edit mode
6.5 years ago
hsbinf
▴
30
Hi,
I used Varscan somatic to get somatic variant calls. In some SNPs, the genotype for normal sample is 1/1 and the alt read count in the normal is 0. I am wondering if this is an acceptable call? Is there a reason I'm missing why Varscan assigned 1/1 to normal despite having 0 alternate reads?
chrom position ref alt normal_genotype alt_genotype alt_count_normal alt_count_tumor total_count_normal total_count_tumor vaf_normal vaf_tumor
chr3 20082072 G A 1/1 0/1 0 2 26 10 0 0.2
Thanks!
The header line you posted has 10 fields, but line 2 has 12. Please post the full command line, and the version of VarScan (I assume you used VarScan2, latest version right)? Did you do any postprocessing of the primary call set?
Varscan v2.3.9
Command: java -jar Varscan.jar somatic normal_pileup tumor_pileup basename --output-vcf 1
Reference file: hg19
Ok, latest version is 2.4.3. I posted my suggested workflow using VarScan recently today here, you should be able to adapt it one-to-one. I suggest you re-run it on your sample with the latest version of the program (maybe on a subset of your BAM including the region you posted). Note that you have to post-process your initial VCF with
processSomatic
in order to split somatic from germline calls, as indicated in my linked post. Have a look at it, and feel free to come back in case of questions.Thanks! Will try that.
Hi this is first line of the result after using Varscan v2.4.3 and doing processSomatic. The normal genotype is still 1/1 even though VAF is 0%. I'm just having trouble understanding why its 1/1 as 1/1 would indicate homozygous alternate but we see 0 alternate reads in the normal.
Also this line is from the somatic high confidence file.