This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Samtools Vcf Output

What does X in the alt column stand for when I use,

-A keep all possible alternate alleles at variant sites, in the bcftools view command for calling snp's.

 chr2 642 . C T,X 85.6 . DP=306;VDB=0.0002;AF1=0.02819;AC1=1;DP4=144,144,3,2;MQ=37;FQ=85.7;PV4=1,0.19,1,1 
GT:PL:DP:SP:GQ  0/0:0,36,255,36,255,255:12:0:48  0/0:0,6,72,6,72,72:2:0:18  0/0:0,72,255,72,255,255:24:0:84  
**0/1:121,0,129,136,144,255:10:0:99**  0/0:0,57,255,57,255,255:19:0:69  0/0:0,48,255,48,255,255:16:0:60 0/0:0,54,255,54,255,255:18:0:66    0/0:0,21,187,21,187,187:7:0:33 0/0:0,66,255,66,255,255:22:0:78  0/0:0,30,234,30,234,234:10:0:42  0/0:0,51,255,51,255,255:17:0:63  0/0:0,36,246,36,246,246:12:0:48    0/0:0,69,255,69,255,255:23:0:81  0/0:0,45,255,45,255,255:15:0:5

Also I see that in the DP4 values, there are only 5 reads supporting the alternate allele.When we apply depth filter, as far as my knowledge goes, we filter by total coverage i.e.DP.Looking at DP4, how can this be called a good SNP (snp quality of 85.6)

variant samtools vcf

2 answers

85.6 is not a great score. With multiple samples in one vcf, a really good SNP will have a score of 999.

Thanks swbarnes.I have few more questions.I am actually looking into filtering snp's.

1.-d and D, this filters snps on read depth.We would be more interested in reads supporting the non-ref allele(last two values of DP4).I don't get the use of total read depth.

2.I filtered the reads by choosing a MAPQ of 20 in samtools mpileup step.In that case, I am expecting a RMS MQ in VCF >=20.I have seen some papers filter SNP's again by RMS MQ <25.Any idea why this is done?

I feel that these questions are relevant to my previous question.I will be happy to start a new post with these questions.Any help is appreciated.

I am not sure what does it mean. But here is the reference:

http://samtools.sourceforge.net/mpileup.shtml

In the BCF generated by SAMtools, an non-ref base 'X' represents an base has not been seen from the alignment data. Such a base is necessary to evaluate the probability of missing a non-ref allele due to sampling fluctuation.

Log in to answer this question.