Heritability - A Paradox of Quantitative Genetics

Heritability is the foundation of quantitative genetics, yet the way it is defined often conflicts with how it is estimated. On one hand, heritability represents the proportion of phenotypic variance explained by genetic variance in the population from which data are sampled. On the other hand, heritability estimated with family and pedigree data represents the heritability in a hypothetical base population where all individuals are assumed to be independent and non-inbred. The discrepancy may not be obvious to many people in the quantitative genetics community. This study demonstrates the discrepancy and, more importantly, introduces a pedigree sparsity coefficient (PSC) to correct the genetic variance/heritability from the base population to the current population. For very dense pedigree data, the correction may allow breeders to better understand the genetic basis of the trait of the current population and predict genetic gain for selection within the current population. The theory and method have been validated with simulated data and data collected from a long-term selection experiment in house mice. The PSC also applies to genomic heritability, where the pedigree relationship is replaced by a genomic relationship matrix calculated from genome-wide markers.