Rheumatoid arthritis (RA) affects millions world-wide. While anti-TNF treatment is widelyused to reduce disease progression, treatment fails inBone-third of patients. No biomarkercurrently exists that identifies non-responders before treatment. A rigorous community-based assessment of the utility of SNP data for predicting anti-TNF treatment efficacy in RApatients was performed in the context of a DREAM Challenge (http://www.synapse.org/RA_Challenge). An open challenge framework enabled the comparative evaluation of predictionsdeveloped by 73 research groups using the most comprehensive available data and covering awide range of state-of-the-art modelling methodologies. Despite a significant geneticheritability estimate of treatment non-response trait (h2¼0.18,Pvalue¼0.02), nosignificant genetic contribution to prediction accuracy is observed. Results formally confirmthe expectations of the rheumatology community that SNP information does not significantlyimprove predictive performance relative to standard clinical traits, thereby justifying arefocusing of future efforts on collection of other data

G.P. is partially supported by NIH grant# R01GM114434 and an IBM faculty award. E.S. is funded by NIH R01GM105857. The Corrona CERTAIN study is sponsored by Corrona, LLC with support from the Agency for Healthcare Research and Quality (R01HS018517). The majority of funding for the planning and implementation of/nCERTAIN was derived from Genentech, with additional support for substudies from Eli Lilly, Momenta harmaceuticals and Pfizer. CERTAIN investigators also receive support from the National Institute of Health (JRC AR053351, JDG AR054 412).