SNP coordinates and 1KG

Question

[muraved]

I have some genotype data which gives me some problems, and I suspect that the genome build is not what it is supposed to be. I have a bim file with the following line:

1 rs2465136 0 980280 G A

According to dbSNP (https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=2465136), this is a SNP on chr1, but the coordinate matches no genome build (1055037 or 990417, instead of 980280). However, a little further down that page, that number occurs in a 1KG handle (1000GENOMES|CEU.trio.12.15.2008_268_chr1_980280). I'm sure I'm missing something obvious here, but could someone explain to me what is going on?

Answer 1

[finswimmer]

Hello,

the position given above belongs to NCBI36 (hg18).

The way I've found it, was to use UCSC liftOver.

fin swimmer

[muraved]

Thanks. Didn't know dbSNP did not provide hg18 coordinates.

[finswimmer]

You're welcome.

hg18 is that old. It releases 2006. hg19 releases 2009. Since 2013 we have hg38. There is no need to support that old reference genomes.

BTW: Does anybody know why gnomAD was build based on h19 and not hg38?

[muraved]

Well, that's a matter of perspective ;-) If one works with older data, like myself, there's a need for that as demonstrated by this question ;-) Would be nice to get ALL available information for a given Snp.

[Kevin Blighe]

muraved, unrelated but maybe this will be of interest: A: Alternate nucleotide is more frequent than reference nucleotide. OMG I'm dizzy.

I very rarely come across datasets/programs that require or are based on hg18, but they obviously do exist. I think that I even saw hg16 one time. As long as the build is clearly stated, you have covered yourself. Generally, though, we should be moving to hg38 where possible.