What Depth Of Coverage For Whole Genome Snp Calling?

Question

[Rubal]

Hey all,

I'm about to embark on a whole genome sequencing project and I'd like to hear peoples' opinions on the coverage you need to be able to start calling SNPs, atleast with some reasonable confidence. I imagine you need ATLEAST 2x, but is there a gold standard?

Best

Rubal

[Sean Davis]

What technology? Tumor/normal? How many samples?

Answer 1

[Sean Davis]

Have a look at this study:

Ajay SS, Parker SC, Ozel Abaan H, Fuentes Fajardo KV, Margulies EH. Accurate and comprehensive sequencing of personal genomes. Genome Res. 2011 Sep;21(9):1498-505. Epub 2011 Jul 19. PubMed PMID: 21771779.

http://www.ncbi.nlm.nih.gov/pubmed/21771779

[Sean Davis]

The cost of false-positive and false-negative results can be quite high, also.

[brentp]

+1 and a nice discussion here.

[Rubal]

now that's gonna be expensive...

Answer 2

[Jfk]

yeah just read that paper, 100+x coverage for accurate calling ...

[Rubal]

yes, i am conducting a population study so by sequencing multiple individuals at lower coverage it approaches 100X coverage if you treat the many individuals as one individual representative of the population. This is applicable to me because I am most interested in high frequency snps. I accept it will be harder to call low-freq snps this way. The 100x coverage for personal genomics makes more sense, in that case you are only interested in a highly accurate genome for 1 individual, so its more affordable.