This is from 4 years ago (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3471395/ http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3471395/) and didn't even involve 23a

This is from 4 years ago (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3471395/ http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3471395/) and didn't even involve 23andMe (way to go, 23andMe, you've sure been making good use of all that donated phenotype info...). A more relevant link might be http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4650984/ http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4650984/ ; the GCTAs show that the common SNPs examined do explain a decent fraction, but it's increasingly looking like rare variants and novel mutations may explain the rest.