[ensembl-dev] Large amount of (Currently?) retired rs# used as primary ID for STAR-gtype SNPs

[Ma, Man Chun John]

Hi,

While I was comparing data between Ensembl and dbSNP for the STAR SNPs
(ie rat SNPs that have among others a ENSEMBL:STAR-gtype origin), I
noticed that while many of them carry a rs# as their primary ID, dbSNP
lists those rs# as retired.

For example, the SNP at 17:56086385 was originally submitted by
STAR-gtype as ENSRNOSNP2801811 and merged as the Celera SNP
ENSRNOSNP991796. It is now listed in Ensembl 59 using the rs# rs65269783
as the primary ID.

However, when I searched dbSNP (NOT Entrez SNP) using this rs#, I
received this error message:

"rs65269783 was deleted on Dec 21, 2009 because its subsnp_id was
deleted:

The snp_id was linked to the following subsnp_id: ss91418351"

The relevant subSNP ss91418351, however, was listed to be uploaded on
Dec 28, 2009 as part of RAT_COMPUTATIONAL_CELERA.

On the other hand, STAR SNPs that still have a valid rs# were listed to
have the only subSNP updated on Feb 26, 2008 (For example, see
rs65700951, aka ENSRNOSNP2801819) as part of RAT_COMPUTATIONAL_STAR.

I guess what happened was when Ensembl uploaded the
RAT_COMPUTATIONAL_CELERA (aka ENSEMBL:celera) set last year, they also
retired all positionally overlapping ss# from RAT_COMPUTATIONAL_STAR
(aka ENSEMBL:STAR-gtype). As dbSNP would retire a rs# when all their
underlying ss# are deleted, all those SNPs are currently invalid.
However, if the precedent of PHARMGKB (SNP set uploaded 2003, retracted
for correction 2006 and resubmitted 2007) holds true, the rs# *may* be
re-activated when dbSNP b131-rat is released. 

Of course, being a novice in these matters, I'm not sure if that's what
is happening now-- and, in the meantime, how are we going to deal with
those rs# that appears invalid...?

John MC Ma
Graduate Assistant
Kwitek Lab
Department of Internal Medicine
3125E MERF
375 Newton Road
Iowa City IA 52242