Program Planner Home Print Page Share Page

Session 204 - Complex Traits and Polygenic Disorders - Posters

2561 - A meta-analysis of height in 4.1 million European-ancestry individuals identifies ~10,000 SNPs accounting for nearly all heritability attributable to common variants

October 26, 2020, 6:00 AM - 11:59 PM Virtual
Authors
L. Yengo1, A. R. Wood2, S. Vedantam3, E. Marouli4, J. Sidorenko5, S. Sakaue6, S. Raghavan7, G. Lettre8, Y. Okada9, J. N. Hirschhorn10, P. M. Visscher11, Genetic Investigation of Anthropometric Traits (GIANT) consortium;
1Program in Complex Traits Genomics, Institute for Molecular Bioscience, Brisbane, Australia, 2Genetics of Complex Traits, College of Medicine and Health, University of Exeter, Exeter, Devon, UK, 3Endocrinology, Boston Children's Hospital, Boston, MA, USA, 4William Harvey Research Institute, Barts and The London School of Medicine and Dentistry Queen Mary University of London, London, UK, 5Institute for Molecular Bioscience, Brisbane, Australia, 6Osaka University, Suita, Japan, 7US Department of Veterans Affairs, Eastern Colorado Healthcare System, Aurora, CO, USA, 8Montreal Heart Inst, Montreal, QC, Canada, 9Department of Statistical Genetics, Osaka University, Suita, Osaka, Japan, 10Department of Pediatrics, Boston Children's Hospital, Boston, MA, USA, 11Institute for Molecular Bioscience, Univ Queensland, Brisbane, Australia.
Disclosures
 L. Yengo: None.
Abstract
Human height is a highly heritable and polygenic trait, with an estimated SNP-based heritability (h2SNP) from common SNPs between 40% and 50% in individuals of European ancestry. Genome-wide significant (GWS) SNPs identified by previous GWAS account for 25% of phenotypic variation, that is less than 50% of h2SNP. Here, as part of a study across 5 major ancestries, we perform the largest GWAS of height to date, in 4.1 million individuals of European ancestry. We identify ~10,000 conditionally independent associations, which altogether explains 40% of height variance (i.e. between 75% and 100% of h2SNP). We find that 1,547/1,703 (i.e. ~90%) of European-ancestry linkage disequilibrium (LD) blocks across the genome contain at least one association and that the genomic distribution of GWS SNPs is not random, such that a few genomic loci contain >20 independent associations within 100 kb of one another, consistent with the presence of multiple causal variants. We find a ~3.2-fold enrichment (P<0.001) of genes with pathogenic mutations causing extreme height or abnormal skeletal growth syndromes near genomic loci with a high density of GWS SNPs. A polygenic score (PGS) of height calculated from all GWS SNPs has a within-ancestry prediction accuracy R2 of 40% (S.E. 3.1%) and combining that PGS with the average height of parents reaches an unprecedented accuracy of 54% (S.E. 3.0%), consistent with theoretical expectation. Altogether, our results show that GWAS of height in individuals of European ancestry approaches saturation for common variants and suggest, when the experimental sample size is sufficiently large, that genetic variation can be resolved to the contribution of individual variants.