SNP coordinate systems in UCSC LiftOver (hg19), NCBI Remap (GRCh37(hg19)) and dbSNP (GRCh38(hg38))

Question

[curious]

Hi,

I'm trying to understand the difference between the coordinate systems of different software: UCSC's LiftOver, dbSNP and NCBI Remap.

I've a list of SNPs with genomic coordinates in hg18 (based on HapMap3 data) that I'm converting to hg19 using liftOver. I wanted to verify if all the coordinates and SNPs reported by liftOver are valid and hence submitted the list of SNPs to dbSNP on browser. But dbSNP returns coordinates in GrCh38 (hg38) which is their latest build. Hence, I used NCBI's Remap to map the coordinates to GrCh37 (hg19).

But the coordinates I've from liftOver and Remap are off by 1. For one instance,

Remap reports chr10 114748496 114748497 rs4074720 and

liftOver returns chr10 114748497 114748498 rs4074720

Is it because dbSNP's coordinate system is 0-based and liftOver is 1-based? I've read posts that mention bed (which was the format I requested from dbSNP) files are 0-based as opposed to other files which can have 1-based coordinate system.

Thanks for your help!

Answer 1

[WouterDeCoster]

Typical off by one error :) You can read some about it (from UCSC) here: https://genome.ucsc.edu/FAQ/FAQtracks.html#tracks1

[curious]

thanks for the reply