Abstract 19854: Whole Genome Sequences and Plasma Lipids in 2,255 Participants
Introduction: Plasma lipids are heritable risk factors for coronary heart disease. High-coverage whole genome sequences allow enumeration of DNA sequence variants of all types (point mutations, structural variants) in different genomic locations (coding, non-coding), and across full range of frequencies (common, rare).
Hypothesis: Common and rare DNA sequence variants in coding and non-coding sequence associate with plasma lipids.
Methods: We performed whole genome sequencing to 30X in 2,255 participants from the Estonian Biobank. We tested the association of plasma lipids with DNA sequence variants using three models: 1) common variant association analysis; 2) rare variant association analysis in coding sequence (allele frequency < 1%, collapsed by gene); and 3) rare variant association analysis in non-coding sequence limited to 301 genes at loci established by genome-wide association studies.
Results: Across 2,255 participants, whole genomes sequencing identified 32.4 million DNA sequence variants (29.8 million single nucleotide polymorphisms, 2.6 million insertions-deletions). Common variant analyses replicated established associations at APOE with LDL cholesterol, CETP with HDL cholesterol, LPA with lipoprotein(a) cholesterol, and APOA5 and LPL with triglycerides. Rare variant analysis in coding sequence replicated associations at Mendelian genes, such as LDLR and APOB for LDL cholesterol. Using chromHMM annotations from ENCODE and Roadmap in a hepatocyte cell line, we identified rare variants at enhancers within 20 kilobases of the transcription start site of the 301 genes. When grouping non-coding variants by gene, the following genes demonstrated associations with nominal p-value <0.05: NF1, SBNO1, APOA5, and APOE for LDL-C; SERINC3, ABCG5, NBEAL2, LRP4, and ANGPTL3 for HDL-C; and PLCB3, HIST1H1B, APOB, SERBP1, FADS1, and SLC22A18AS for triglycerides. Replication of these results is on-going in an additional 8,000 participants who have been whole genome sequenced.
Conclusions: Whole genome sequences allow for the analysis and discovery of rare, regulatory variants associated with plasma lipids.
Author Disclosures: P. Natarajan: None. S.M. Sekavat: None. A. Ganna: None. A. Metspalu: None. B. Neale: None. T. Esko: None. S. Kathiresan: Employment; Modest; Regeneron, Merck, Celera, Novartis, Bristol-Myers Squibb, Sanofi, AstraZeneca, Alnylam, Eli Lilly, Leerink, Catabasis,. Research Grant; Significant; Regeneron, Bayer. Consultant/Advisory Board; Modest; Regeneron.
- © 2016 by American Heart Association, Inc.