Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement
The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.
Gene Function Relationships to Discover Novel Rheumatoid Arthritis Risk Loci
Raychaudhuri1, S., Remmers2, E.F., Eyre3, S., Xie4, G., Alfredsson5, L., Coblyn1, J., Criswell6, Lindsey A.
Brigham & Women's Hospital, Boston, MA
Academic Medical Center/University of Amsterdam, Amsterdam, Netherlands
GENRA Consortium, Amsterdam, Netherlands
Brigham and Women's Hospital. Boston, MA
Leiden University Medical Centre, Leiden, Netherlands
Roche Diagnostics, Pleasanton, CA
University of Toronto
arc Epidemiology Unit, Manchester, United Kingdom
Broad Institute, Cambridge, MA
The University of Manchester, Manchester, United Kingdom
University of Toronto, Toronto
Karolinska Institute, Stockholm, Sweden
University of California San Francisco, San Francisco, CA
Leiden University Medical Center, Leiden, Netherlands
Feinstein Insititute Med Rsch, Manhasset, NY
Karolinska Institutet, Stockholm, Sweden
Genetic studies in RA have identified 16 loci containing genes involved in key immunologic processes. We hypothesized that undiscovered RA risk loci likely contain genes functionally related to previously validated gene loci, and that a systematic approach to establishing those functional relationships will effectively help identify novel RA loci from genome-wide association studies (GWAS).
To test this hypothesis, we examined 370 SNPs from 179 independent loci with p<0.001 in a published meta-analysis of three RA GWAS (Raychaudhuri et al, Nat Genet 2008), most of which likely represent spurious associations. We developed and applied GRAIL (Gene Relationships Across Implicated Loci, Raychaudhuri et al PLOS Genet), a computational method that automatically assesses functional relationships between genes by applying statistical text mining to PubMed abstracts. We scored these 179 loci for functional relationships to genes in 16 known RA disease loci. We genotyped representative SNPs from loci with the highest scores in an independent set of 8,096 autoantibody positive RA cases and 11,822 matched controls.
We identified 22 loci with a high degree of functional connectivity as determined by high GRAIL scores. Three of these validate convincingly: CD2/CD58, PRDM1, and CD28 (see Table 1). In addition, four loci replicate at p<0.0023 (=0.05/22), indicating that these almost certainly represent true RA risk loci: PTPRC, TAGAP, TRAF6/RAG1 and FCGR2A. Many of these loci are already associated to other immunologic diseases.
Table 1. 7 of 22 SNPs tested replicate in independent samples (p<0.0023).
|SNP||Meta-Analysis (3,393 Cases; 12,460 Controls)||Replication (8,096 Cases; 11,822 Controls)||Joint|
|rs11586238||1p13.1||CD2, IGSF2, CD58||2.0E-04||1.14||3.0E-06||1.11||0.228||0.253||2.3E-09||1.12|
We present novel RA risk loci, discovered by utilizing pathway-based approaches; these loci could indicate key immunologic processes in pathogenesis.
To cite this abstract, please use the following information:
Raychaudhuri, S., Remmers, E.F., Eyre, S., Xie, G., Alfredsson, L., Coblyn, J., et al; Gene Function Relationships to Discover Novel Rheumatoid Arthritis Risk Loci [abstract]. Arthritis Rheum 2009;60 Suppl 10 :574