A1512
Title: Statistical analysis of the genetic correlations between quantitative traits
Authors: Chen-Hung Kao - Institute of Statistical Science (Taiwan) [presenting]
Abstract: It is well known that phenotypic correlation between two traits is an outcome of interplays among genetic correlation, residual correlation, and heritability. Also, genetic correlations between different quantitative traits are caused by the closely linked and pleiotropic QTL. It shows by example that, for a pair of traits, high (low) phenotypic correlation does not imply high (low) genetic correlation, and vice versa, as the heritability and residual correlation also play roles in affecting the strength of phenotypic correlation. The phenotypic and genetic correlations seem to have an arbitrary relationship, making accurate prediction of each other difficult. Statistical analysis is presented to dissect the genetic correlations between traits by showing that, depending on the effect sizes and linkage parameters, multigenic traits controlled by tightly linked and pleiotropic QTL do not necessarily show significant genetic correlations with each other, because linkages between different QTL may make different levels of (positive or negative) contribution and may combine to produce no genetic correlation, or very weak to very strong genetic correlations. Analytical and numerical examples are presented.
