Abstract
To establish a simple and effective method for personalized medication, we attempted to predict drug therapy efficacy with Support Vector Machine (SVM) based on the gene expressions of peripheral blood cells from 73 anonymous rheumatoid arthritis (RA) patients. For this purpose, the patients treated with anti-TNF-α antibodies were evaluated drug therapy efficacy assessment based on the American College of Rheumatology criteria. We categorized the patients into two groups: responders against the drug and non-responders. The genes with significant differences between responders and non-responders before medication were chosen by the Significance Analysis of Microarray (SAM) method and patients were discriminated by SVM by using the expression of those genes, so that we identified the essential 9 genes to predict drug efficacy. Next, we investigated the genes significantly changed after medication in responders. Consequently, our results support the hypothesis that suppression of osteoclasts differentiation by IFN-γ contribute to symptomatic improvement of RA patients, and suggest that the result of SVM is consistent with that from experiment.