SNP Calling Parameter Settings

Question

[mcclintock]

Hi everyone,

I am using family trios data to run GATK's SNP calling pipeline. The HaplotypeCaller gave me the father's SNPs file, which I have recalibrated using VariantRecalibrator. But the final VCF file still has 3 million “PASS” records. Actually, one human has no chance to carry so many SNPs.

Any advice for adjusting the parameters?

The parameters I used refer to the literature below.

Roazen, D., Thibault, J., Banks, E., Garimella, K., Altshuler, D., Gabriel, S. and DePristo, M. (2013). From FastQ Data to High-Confidence Variant Calls: The Genome Analysis Toolkit Best Practices Pipeline. Current Protocols in Bioinformatics, pp.11.10.1-11.10.33.

Thanks.

[Fabio Marroni]

Actually, one human has no chance to carry so many SNPs.

Why do you think so? What is the expected number of SNPs for one human subject on average?

[mcclintock]

A little more than 2 million is appropriate, as far as I know.