Skip to main content
Fig. 4 | Arthritis Research & Therapy

Fig. 4

From: Identification of baseline gene expression signatures predicting therapeutic responses to three biologic agents in rheumatoid arthritis: a retrospective observational study

Fig. 4

An overview of classification of all 209 samples using 3 signature scores. All 209 samples could be classified into 8 groups, based on the binary variables derived from the signature scores of the three gene sets (i.e., inflammasome-, specific CD19-, and specific CD56-related). The threshold for binary call was determined using ROC curve analysis. Top panel: Heat map of 209 samples based on expression patterns of the three core gene sets. The heat map was performed based on relative expression levels (z-scores) of core genes using TIGR MultiExperiment Viewer Software (http://www.tm4.org/). Middle panel: Prediction results using signature scores. Patient outcomes were predicted using signature score and grouped as remission (REM) and nonremission (NON-REM) as indicated in green and red, respectively. Bottom panel: Actual remission status based on CDAI at 6 months of biologic therapy. In “Individual outcome,” green and red circles represent actual individual cases achieving or not achieving remission, respectively. In “Rate of NON-REM,” the actual number of cases and the nonremission rates are indicated for all eight groups. REM patients with CDAI remission at 6 months of biologic therapy, NON-REM patients without CDAI remission at 6 months of biologic therapy, CDAI clinical disease activity index, IFX infliximab, TCZ tocilizumab, ABT abatacept 

Back to article page