Best practices and proper interpretation of SNP-heritability estimates

Multiple methods have been developed to estimate narrow-sense heritability using single nucleotide polymorphisms in unrelated individuals (“SNP-heritability”). However, a comprehensive evaluation of these methods has not... [ view full abstract ]

Multiple methods have been developed to estimate narrow-sense heritability using single nucleotide polymorphisms in unrelated individuals (“SNP-heritability”). However, a comprehensive evaluation of these methods has not yet been performed, leading to confusion and discrepancy in the literature. We present the most thorough and realistic comparison of these methods to date using thousands of real whole genome sequences to simulate genotypic and phenotypic data under various scenarios. We show that SNP-heritability estimates can be highly sensitive to assumptions about the frequencies, effect sizes, and levels of linkage disequilibrium of underlying causal variants. We introduce a novel procedure (LDMS-I) that leads to estimates that are less sensitive to these assumptions across a wide range of genetic architectures and possible confounding factors. Our findings provide guidance for best practices and proper interpretation of published estimates.