Estimating heritability using genomic data

Heritability (h(2)) represents the potential for short-term response of a quantitative trait to selection. Unfortunately, estimating h(2) through traditional crossing experiments is not practical for many species, and even for those in which mating can be manipulated, it may not be possible to assay them in ecologically relevant environments. We evaluated an approach, GCTA, that uses relatedness estimated from genomic data to estimate the proportion of phenotypic variance due to genotyped SNPs, which can be used to infer h(2). Using phenotypic and genotypic data from eight replicates of experimentally grown plants of the annual legume Medicago truncatula, we examined how h(2) estimates from GCTA (h(GCTA)(2)) related to traditional estimates of heritability (clonal repeatability for these inbred lines). Further, we examined how h(GCTA)(2) estimates were affected by SNP number, minor allele frequency, the number of individuals assayed and the exclusion of causative SNPs. We found that the average h(GCTA)(2) estimates for each trait made with the full data set (> 5 million SNPs, 200 individuals) were strongly correlated (r = 0.99) with estimates of clonal repeatability. However, this result masks considerable variation among replicate estimates of h(GCTA)(2), even in relatively uniform greenhouse conditions. h(GCTA)(2) estimates with 250000 and 25000 SNPs were very similar to those obtained with > 5 million SNPs, but with 2500 SNPs, h(GCTA)(2) were lower and had higher variance than those with >= 25 k SNPs. h(GCTA)(2) estimates were slightly lower when only common SNPs were used. Excluding putatively causative SNPs had little effect on the estimates of h(GCTA)(2), suggesting that genotyping putatively causative SNPs is not necessary to obtain accurate estimates of h(2). The number of accessions sampled had the greatest effect on h(GCTA)(2) estimates, and variance greatly increased as fewer accessions were included. With only 50 accessions sampled, the range of h(GCTA)(2) ranged from 0 to 1 for all traits. These results indicate that the GCTA method may be useful for estimating h(2) using data sets of a size that are available from reduced-representation genotyping but that hundreds of individuals may need to be sampled to obtain robust estimates of h(2).