Choosing SNPs

[bounlu]

How to choose SNPs as the input for alleleCount? More specifically;

1. How many SNPs are sufficient throughout the genome?
2. How should the SNPs be distributed along the genome? i.e. in terms of window size, genomic region, ... ?

[AndyMenzies]

All AlleleCount does is to provide the base composition at the locations you specified. How many locations you need is entirely down to your use case. For example

• We use 4 specific targeted snps for a human sample gender check
• We use a set of 96 highly polymorphic snps for a quick sample identity check
• We use a panel of 2 million evenly spaced snps to feed into ASCAT, a copy number assessment algorithm.

It depends on your end goal.

[bounlu]

Thank you very much for the helpful information.

Could you please provide me the list of SNPs in hg38 you mentioned?

[AndyMenzies]

We don't have many of those snp sets publicly available, we do have a reference set of the ASCAT snps available for GRCh37, they can be found here

ftp://ftp.sanger.ac.uk/pub/cancer/support-files/CPIB/ascatNgs/Human/GRCh37/legacy