Identification of correlated genetic variants jointly associated with rheumatoid arthritis using ridge regression

Abstract Using the North American Rheumatoid Arthritis Consortium genome-wide association dataset, we applied ridged, multiple least-squares regression to identify genetic variants with apparent unique contributions to variation of anti-cyclic citrullinated peptide (anti-CCP), a newly identified clinical risk factor for development of rheumatoid arthritis. Within a 2.7-Mbp region on chromosome 6 around the well studied HLA-DRB1 locus, ridge regression identified a single-nucleotide polymorphism that was associated with anti-CCP variation when including the additive effects of other single-nucleotide polymorphisms in a multivariable analysis, but that showed only a weak direct association with anti-CCP. This suggests that multivariable methods can be used to identify potentially relevant genetic variants in regions of interest that would be difficult to detect based on direct associations.http://deepblue.lib.umich.edu/bitstream/2027.42/78271/1/1753-6561-3-S7-S67.xmlhttp://deepblue.lib.umich.edu/bitstream/2027.42/78271/2/1753-6561-3-S7-S67.pdfPeer Reviewe