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


From SNPedia

In the interest of uniformity of entries, I'm wondering if there are preferences for scientific notation to use. Also describing statistical signficance is fraught with danger, especially if taking the research papers at face value (there has been considerable discussion about the definition of statistical significance for such studies in recent years). Perhaps the P-values should be let to speak for themselves, but unfortunately study size and arrangement such as candidate SNP selection and verification have an effect. On reporting Odds Ratio, for example the NHGRI GWAS database used for much of SNPedia content always flips the allele so it's reporting incrased risk. Many if not most studies itself have adopted the convention of reporting OR for minor allele, whether it happens to be increased or decreased risk; this somewhat alleviates the concern for ambigous flips, but can still be problematic when allele frequencies are close or vary between populations. Of course, flipping the OR and CI for the risk allele involve a calculation which could potentially go wrong, and makes the numbers harder to verify against original study. So I'm not sure which style would be clearer and preferred. --Donwulff (talk) 08:28, 17 September 2014 (UTC)

Neither am I. Do what seems best to you. The wiki doesn't need perfection on the first try. Steady small improvements work best. --- cariaso 12:47, 17 September 2014 (UTC)