mpileup-varscan

[bioman1]

Hi all,
I have aligned illumina reads to the reference genome using bowtie and then created mpileup file using sam tools. I have used varscan to call variants (SNPS) using mpileup2snp with min coverage 20, min reads2 4, min avg qual 20 and p-value 0.1 settings.I would like to determine heterozygosity & homozygosity.

I get output from the varscan for some of the contigs repeated more than once showing showing different p-values, var frequency , strand filter, sample homo & sam hetero. I have counted the snps for each contig which shows strand filter-Pass. But I am confused how can i get p-value, var freq for each contig (if it is repeated more than twice)

Is there any way filter the result in varscan to get values for each unique contig?

[shyam_la]

http://seqanswers.com/forums/showthread.php?t=17079

Check this out. It won't answer your question precisely, but will give a better understanding of the parameters in varscan..
Did you run the mpileup with the -B parameter? Because if you didn't VarScan is bound to give faulty results..

[bioman1]

mpileup2snp

Thanks you..I will try mpileup with -B option..
I also get different results when I tried mpileup2snp with VCF format turned on/off in the options.

I get more number of SNPs predicted when I put VCF format output turned on and get lower number of SNPs when I turned off VCF format off. Do u you know why varscan shows different results?

[shyam_la]

Nope.. My experience with VarScan is very limited. Sorry.

[bioman1]

Thanks a lot..now it is fixed.

Now I get same no of SNPS predicted even when VCF option turned on/off in Varscan, when I create samtools mpileup creation with -B option.

I hope this helps some one

[dkoboldt]

Bioman1,

Thanks for your post, and glad you got it straightened out!

~Dan Koboldt