RT Journal Article
SR Electronic
T1 Assortative Mating Biases Marker-based Heritability Estimators
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 2021.03.18.436091
DO 10.1101/2021.03.18.436091
A1 Border, Richard
A1 O’Rourke, Sean
A1 de Candia, Teresa
A1 Goddard, Michael E.
A1 Visscher, Peter M.
A1 Yengo, Loic
A1 Jones, Mathew
A1 Keller, Matthew C.
YR 2021
UL http://biorxiv.org/content/early/2021/03/20/2021.03.18.436091.abstract
AB Many complex traits are subject to assortative mating (AM), with recent molecular genetic findings confirming longstanding theoretical predictions that AM alters genetic architecture by inducing long range dependence across causal variants. However, all marker-based heritability estimators assume mating is random. We provide mathematical and simulation-based evidence demonstrating that both method-of-moments estimators and likelihood-based estimators produce biased estimates in the presence of AM and that common approaches to account for population structure fail to mitigate this bias. Then, examining height and educational attainment in the UK Biobank, we demonstrate that these biases affect real world traits. Finally, we derive corrected heritability estimators for traits under equilibrium AM.Competing Interest StatementThe authors have declared no competing interest.