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


From SNPedia

I think I found an error in the Promethease output. I was looking at the eye color snps under topics and specifically happened to notice rs2240203. The main text from Promethease states: "89.4% rs2240203(A;A)   usually brown eye color". Based on the text associated with this SNP, someone who is homozygous for A/A at SNP rs2240203 should usually have blue eyes, not brown eyes as the output reports. The same error appears to be present for SNPs rs12593929 and rs1129038 as well.

I've seen this plenty of times and just assumed the statement was true for the h-1 haplotype, but not the haplotype other than the one a particular person has. However when I spot check User:Lilly Mendel I see the exact same 3 snps you've reported above, so it does seem notable. In general there hapmap frequencies are such that I'm inclined to at least blank their genotype summaries, since haplotype specific text belongs on a genoset not a genotype but perhaps there is instead a simpler error as you suggest. Should rs2240203 rs12593929 and rs1129038 have their genotype summaries flipped? cariaso 23:54, 26 July 2011 (UTC)
There are about 13 SNPs associated with this haplotype and the other 10 all have essentially the reverse statement and logic as the statements for these three, which is why I think this is a mistake in the genotype summaries of these three specific SNPs. Whether they should be flipped or the haplotype text should be removed from the specific SNPs is beyond my current understanding of the way the site is organized. msr 26 July 2011
Agreed. I can verify from my genotype (hazel eyes, a blue-eyed mother) and my husband (blue eyes) that the predictions for all three alleles are backwards. --Kay, 1 August 2011
good enough for me, I've made the changes. "Given enough gs237s, all bugs are shallow" - source. cariaso 23:13, 1 August 2011 (UTC)