Use of low read depth whole genome sequence data to examine the genomic architecture of commonly measured lipid sub-fractions: the UK10K study. J. Huang1, J. Min2, V. Iotchkova1, M. Mangino1, A. Gaye1, M. Kleber1, G. Malerba1, M. Cocca1, T. Michela1, I. Tachmazidou1, H. Chheda1, A. Manning1, A. Wood1, R. Scott1, T. Gaunt1, W. Zhang1, F. Rivadeneira1, N. Soranzo1, N. Timpson2, UK10K Consortium Cohorts Group 1) Human Genetics, Wellcome Trust Sanger Institute, Hinxton, UK; 2) MRC Integrative Epidemiology Unit, University of Bristol, UK.

   Lipid levels are highly heritable risk factors for coronary artery disease and vascular outcomes, and augmentation of their levels therapeutically has well characterized impacts on risk profile. Discovery of inherited variation influencing lipid levels has relevance for treatment of cardiovascular disease through the identification of novel therapeutic targets and the assessment of the causal impact of specific profiles. As part of the UK10K Consortium Cohorts arm we undertook low read depth whole genome sequencing in individuals from two deeply phenotyped British cohorts, the St Thomas Twin Registry (TwinsUK) and the Avon Longitudinal Study of Parents and Children (ALSPAC). In this lipid focused initiative, we analyzed the cohorts dataset with the following three aims: (i) to discover additional variants of low and intermediate frequency that are poorly represented in genome-wide SNP arrays; (ii) to characterize the contribution of rare, deleterious variants to population level trait variance, and to rank human genes according to the presence of these variants; (iii) to systematically evaluate the contribution of low-frequency and rare variants in non-coding regions of the genome. We further combined the UK10K data with meta-analysis of 18 additional cohorts (N=~ 25,000) with SNP array data imputed to the combined UK10K+1000 Genomes Project haplotype reference panel. Overall, we describe 21 novel independent genome-wide significant loci (defined as P < 5x10-8). Of these, 18 map to known GWAS signals, including several low-frequency variants (MAF <5%) near PCSK9, LPL, APOC3|APOA4|APOA1, PCSK7, CETP, LIPG, LDLR and APOE. We further use Bayesian approaches to fine-map genetic association signals at known loci, and explore possible functional consequences of putative causative variant through integration with functional annotation maps of the human genome. Finally, we report association and replication results of rare variant tests (MAF <1%) from sequence kernel based association testing at both exome-wide and genome-wide scale. Overall, our results demonstrate the value for low-coverage whole genome sequencing and shed lights on the role of rare genetic variants on serum lipids levels.

You may contact the first author (during and after the meeting) at