Purchase this article with an account.
Xiaowei Zhan, David Larson, Robert Fulton, Chaolong Wang, Dwight Stambolian, Emily Chew, Elaine Mardis, Anand Swaroop, Goncalo Abecasis, ; Targeted Sequencing, Augmented with Public Resources, Identifies a Rare C3 Allele Associated with Large Risk of Age-related Macular Degeneration. Invest. Ophthalmol. Vis. Sci. 2013;54(15):6177.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Macular degeneration is one of the most common causes of incurable blindness. Common alleles in >19 loci have now been associated with disease. We set out to investigate whether rare variants in the same loci were also associated with disease risk and to compare the relative effect sizes of common and rare variants.
In collaboration with the Genome Sequencing Center at Washington University in St. Louis, we designed a sequencing study focused on 8 of the known AMD risk loci (CFH, ARMS2, C3, C2/CFB, CFI, CETP, LIPC and TIMP3/SYN3) and 2 other candidate regions (LPL and ABCA1). We re-sequenced these regions in 3,124 individuals (2,335 cases and 789 controls) to an average depth of 85x. To augment the number of controls available for analysis, we designed an algorithm to identify previously sequenced samples with good coverage of our regions of interest and similar genetic ancestry. Finally, we investigated association between genetic variation in each locus and risk of disease using both single variant and gene-level burden tests.
Across 967 kb of examined sequence, we discovered 41,202 high quality SNP variants in the 3,124 sequenced individuals (34,346 of these were novel, and not previously described in dbSNP). Among these variants, we focused our attention on 1,800 nonsynonymous, stop and splice variants. We estimated the ancestry of our sequenced samples and of samples from the NHLBI exome sequencing project, identifying an ancestry matched set of 2,268 cases and 2,268 controls. Individuals in this matched set had deep (minimum 10x) coverage of coding regions in the 10 targeted loci. Association analysis identified two strongly associated variants, one in the CFH gene (control frequency = 0.02%, exact P-value = 2.91x10-6, OR = 23.11) and another in the C3 gene (control frequency = 0.40%, exact P-value = 2.73x10-4, OR = 2.68). Replication efforts for these findings are ongoing.
Through targeted sequencing efforts, augmented with publicly available control data, we replicated a previously reported rare variant association in the CFH gene and identify a new rare variant signal in the C3 gene. In both instances, these rare variants are associated with substantially larger odds ratios than common variants in the same regions.
This PDF is available to Subscribers Only