Effect Of Using Reads With Mapping Quality >30 For Snp Detection Instead Of All Reads.

Question

[jack]

What would be the effect of only using reads with mapping quality > 30 for SNP detection instead of all reads?

Answer 1

[Istvan Albert]

It will be a quality vs quantity tradeoff. Depending on the data/experiment/tool it will have a small or large effect.

And that effect will be that the SNPs that are called will bit a little bit more reliable (fewer false positives) but the tool will miss existing SNPs (more false negatives)

[Matt Shirley]

I would also add that it depends on whether you want germline or somatic SNPs. Low allele frequency somatic mutations might be a bit less reliable if they are only supported by low quality base calls.