Have questions? Visit https://www.reddit.com/r/SNPedia


From SNPedia

GS criteria don't identify the major/minor alleles (risk/non-risk) succinctly. This seems like a significant oversight (although it doesn't look like the Criteria syntax has been changed since 2009). ResearcherQ (talk) 14:43, 13 March 2015 (UTC)

please suggest a particular gs# you'd like to point to so that we can be specific as we continue this discussion. --- cariaso 17:11, 13 March 2015 (UTC)
Mike, good afternoon. The day came to convert from a lurker to a participant. Thanks for your quick response (I thought I was making notes only for myself).
As I look at the Criteria syntax, I see only True or False. It looks like I'll need to write my own wrappers (and syntax) to make observations on whether risk alleles are present. I see them often expressed as major/minor, such as rs6656401(G/A) -- where GG is mapped to +/+, GA to +/-, and AA to -/-. I was hoping such a succinct primitive was there already, but I'll just tackle it myself this afternoon; meaning, I'll invent a syntax for myself that probably looks like rs6656401(G/A) -- not a semicolon -- and map that to the various calls to rs6656401(G), etc. gs188 looks like a good example for which I'll make a generalized wrapper.
If there's a list of such syntax built up already, please point me to the right place.
Thanks ResearcherQ (talk) 19:45, 13 March 2015 (UTC)
Welcome. Yes, everything does reduce to True/False. Major/Minor is a useful distinction, but doesn't always capture subtle combinations. Gs140/criteria is an example of exhaustive coverage. It sounds like you have what you need for now, or I do not see a specific question I can answer. --- cariaso 19:59, 13 March 2015 (UTC)
GS140 = yuck! Thanks for watching over things - ResearcherQ (talk) 20:18, 13 March 2015 (UTC)