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


From SNPedia

A user writes: My Promethease report indicates that due to my G/T alleles at rs2241766, I have a 0.6x decreased risk of breast cancer. This was welcome information until I looked at the SNPedia page for more details and found that these alleles are only associated with a reduced risk if another SNP (rs1501299) happens to be TT. The significance is whether or not someone is a high,intermediate, or low signaler of adiponectin. In my case, my alleles indicate that I am a "low signaler," since the other SNP is G/G. This means that -- instead of a 0.6x reduced risk, I have a 6.56x higher risk. Am I misunderstanding something? Or should the reporting for rs2241766 indicate that G/T can indicate EITHER a reduced risk - OR - a greatly increased risk, depending on the status of the other SNP rs1501299 ? While assuming I had a 0.6 reduced risk with this SNP, my grand total risk for breast cancer was still a dismal 13x increased risk. However, with this 6.56x increased risk, it now multiplies out to almost 150x higher than average!

The simple summary of 'reduced risk' does appear to have been true in isolation, but in context of the neighboring SNP there is more to consider. I've removed the simple summary to encourage digging deeper.
The two SNPs are both in the same gene, but one of them is considered to be in the + strand, the other is in the -. This is a bit odd, and likely to make our terminology a bit confusing. The good news, is neither one is an ambiguous_flip so there doesn't seem to be any deeper confusion in the literature.
This pair is reported to influence adiponectin signaling, but that is turn has numerous effect, many related to metabolic syndrome, so the consequences are not particularly breast cancer specific.
The 13x number comes from one paper which was reasonably current when this wiki page was last being edited. Since that time, I dont see any substantial replications, so that number should be taken with a large grain of salt.
I don't see other papers widely declaring that these 2 SNPs are THE causative ones, so there may be more or better snps to consider. If they alone are the best choice, Promethease supports genosets which would allow it to identify your high/med/low signalling status according to the table, but it's not yet clear to me that the table is suitably authoritative to warrant that. --- cariaso 11:11, 12 November 2011 (UTC)