Evaluating regional heritability mapping methods for identifying QTLs in a wild population of Soay sheep

Regional heritability mapping (RHM) is a method that estimates the heritability of genomic segments that may contain both common and rare variants affecting a complex trait. We compared three RHM methods: SNP-RHM, which uses genomic relationship matrices (GRMs) based on SNP genotypes; Hap-RHM, which uses GRMs based on haplotypes; and SNHap-RHM, which uses both SNP-based and haplotype-based GRMs jointly. We applied these methods to data from a wild population of sheep, analysed eleven polygenic morphometric traits and compared the results with previous genome wide association analyses (GWAS). We found that whilst the inclusion of the regional matrix did not explain significant variation for all regions that were associated with trait variation using GWAS, it did for several regions that were not previously associated with trait variation.