Genomic Assortative Mating in Human Marriages

Guang Guo, University of North Carolina at Chapel Hill
Lin Wang, University of North Carolina at Chapel Hill
Hexuan Liu, University of North Carolina at Chapel Hill
Yilan Fu, University of North Carolina at Chapel Hill

Using GWAS data and about 1,000 married couples from the Framingham Heart study, this analysis investigates genomic assortative mating in human marriages. The study calculates (1) a SNP-specific correlation for each of the 320,195 SNPs, averaging 1,023 married couples and (2) a married couple-specific correlation for each couple, averaging 287,295 SNPs. Of the 320,195 SNP correlations, eight have a p-value 5x10-8 or smaller. These SNPs are all positively correlated for married couples, with a range of 0.16-0.27. The SNPs implicated in genes associated with BMI, height, risk behavior, or the human leukocyte antigen system are further examined. For the half of all SNPs having a more positive assortment, married couples average a statistically significant correlation that is 0.001 higher than randomly paired individuals. Our simulation shows that this difference in correlation can be explained by married couples assorting on about 200 SNPs.