Strand info, SNPs

[kostask]

Hello,

I am performing an analysis for SNPs in Galaxy UCSC, using the tools mpileup (to generate pileup from bam files) and Varscan (to generate vcf from pileup files).

Although the programs work fine, I haven't been able to find a way to add the column with + and - at the final vcf file, in order to show on which strand does the reported reference base and alternative base belong.

Is there a step before the analysis or after (etc the annotation) from where I can extract this information?

Kostas

[SylvainL]

Hi,

I'm not sure I understand... Normally, in SNP analysis, the reported SNP is always according to the reference (meaning strand +). Then, if you want to know what is the impact if it concerns a coding region, you have to do the translation yourself (meaning then you have to know on which strand the SNP will have an effect).

Except if you did your SNP calling on transcriptome reference (but then, the strand will be the same as the sequences you used as reference).

s.

[kostask]

Actually I used RNA-seq data and I performed the SNP calling step on human genome version hg19.

No it returns both + and - strand since it usually contains reads corresponding to both orientations.

[SylvainL]

I really don't catch it. SNP calling will tell you if the sequence of your read differs from the reference... If you use the genome as reference, it's the base of the genome (then always strand +) which will be used as reference...

The strand your read is mapping is not important, as the fact you use RNAseq (except that the probability of a SNP in coding regions is smaller). With DNAseq you also have reads mapping on both strands...

If you have an example of a SNP reported on the strand -, I would really be happy to have a look at it.