Abstract 12903: Signatures of Recessive Alleles and Susceptibility to Coronary Artery Disease: Genome-Wide Homozygosity Meta-Analysis
Introduction Previous genome-wide association studies (GWAs) have identified several common variants with an additive impact on coronary artery disease (CAD). However, the collective contribution of these variants to CAD is modest and explains only a small proportion (∼13%) of its overall heritability. One possible explanation for the remainder of missing heritability in CAD are highly penetrant recessive alleles. Runs of homozygosity (ROHs) - long segments (typically >1Mb) of uninterrupted sequences of >100 consecutive homozygous SNPs often mark such rare mutations. We explored whether ROHs are more prevalent in patients with CAD than in CAD-free controls and whether homozygosity mapping may be helpful to identify some of the missing CAD heritability.
Methods and Results Using approximately 2.5 million imputed SNPs from previously conducted GWAs in 9 cohorts of white European ancestry (WTCCC, GerMIFS I-II-III, CCGB, DUKE, OHGS A-C and PennCATH) we identified and analysed homozygosity levels of 10,548 CAD patients and 10,273 CAD-free controls. Each individual had approximately 32 ROHs within their autosomal DNA. These tracts of homozygous SNPs had an average length of 1.36±0.11 Mb and cover on average a total length of 51.3Mb (1.8% of human genome). We identified that whilst a majority of ROHs were specific to one cohort, a total of 15,441 ROHs (32.8% of all identified ROHs) was present in ≥2 populations (overlapping ROHs). The overlapping ROHs were over-represented (odds ratio>1) in CAD patients when compared to controls [8,502 (55%) vs 6,939 (45%)]; binomial test P=1.34x10[[Unable to Display Character: ]]36. This enrichment in CAD was independent of the length or chromosomal location of the ROHs. However, there was no association between CAD and any individual overlapping ROH after correction for multiple testing.
Conclusion The difference in global genetic architecture of overlapping ROHs between those with and without CAD suggests the presence of many recessive alleles with a potential impact on the risk of CAD.
- © 2012 by American Heart Association, Inc.