Different output from pileup and mpileup commands VarScan

Question

[agata88]

Hello,

In last days I was testing Varscan mpileup and pileup commands. Here are commands that I've used:

1) samtools mpileup -m1 -B -F0.0001 -d5000 -f <reference> <bam> > bam.mpileup

A) VarScan.jar pileup2snp input --strand-filter 0 –min-coverage 3 –min-var-freq 0.1

B) VarScan.jar mpileup2snp input --output-vcf --strand-filter 0 –min-coverage 3 –min-var-freq 0.1

Unfortunately I get different number of SNPs detected. For 1A - 93587 and 1B - 53769. I am comparing results to other tool which predicted variants with the number of 8827. I would like to have results including those 8827 snps. With the use of pileup I have 100% covered but with mpileup only 95,70%.

I need to use mpileup command for VCF output which wll be further use for SNPEff Annotation.

Could anyone help me with this issue?

I would really appreciate for help,

Best,

Agata

Answer 1

[agata88]

So since nobody have answered this question I am going to do that:)

It seems that there are some differences in default p-value. When I run second command (B) with p=0.05 (not 0.01 as default) I have all of variants that I need (including 100% of 8827 variants predicted with other tool).

So, the conclusion is to manipulate with the probability threshold to get results more fit with your desire results,

Best,

Agata