Estimating heritability using genomic data

SummaryHeritability ( 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 SNP s, 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 2 GCTA ) related to traditional estimates of heritability (clonal repeatability for these inbred lines). Further, we examined how h 2 GCTA estimates were affected by SNP number, minor allele frequency, the number of individuals assayed and the exclusion of causative SNP s. We found that the average h 2 GCTA estimates for each trait made with the full data set (>5 million SNP s, 200 individuals) were strongly correlated ( r  = 0·99) with estimates of clonal repeatability. However, this result masks considerable variation among replicate estimates of h 2 GCTA , even in relatively uniform greenhouse conditions. h 2 GCTA estimates with 250 000 and 25 000 SNP s were very similar to those obtained with >5 million SNP s, but with 2500 SNP s, h 2 GCTA were lower and had higher variance than those with ≥25 k SNP s. h 2 GCTA estimates were slightly lower when only common SNP s were used. Excluding putatively causative SNP s had little effect on the estimates of h 2 GCTA , suggesting that genotyping putatively causative SNP s is not necessary to obtain accurate estimates of h 2 . The number of accessions sampled had the greatest effect on h 2 GCTA estimates, and variance greatly increased as fewer accessions were included. With only 50 accessions sampled, the range of h 2 GCTA 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 .