Multiple SNPs within few bases

Question

[Chirag Nepal]

Hi there,

Some of the predicted somatic snippers are close, 2/3 SNPs within 10 nt window.

Varscan actually filters such SNPs when using their filtering scripts.

So SNPs that are close to each other, or close to indel, why are such SNPs discarded. If I want to filter such SNPs using some distance, what distance, 10nt or 20nt, what might be appropriate.

Thanks!

[Sam]

Usually, if multiple snps are observed in a short distance e.g. 1 read length, then it is very likely there is an indel in the area and you should try to first perform the local realignment before calling.