Intravenous immunoglobulin (IVIG) is used to treat several autoimmune and inflammatory diseases, but some patients are refractory to IVIG and require alternative treatments. Identifying a biomarker that could segregate IVIG responders from non-responders has been a subject of intense research. Unfortunately, previous transcriptomic studies aimed at addressing IVIG resistance have failed to predict a biomarker that could identify IVIG-nonresponders. Therefore, we used a novel data mining technique on the publicly available transcriptomic data of Kawasaki disease (KD) patients treated with IVIG to identify potential biomarkers of IVIG response. By studying the boolean patterns hidden in the expression profiles of KD patients undergoing IVIG therapy, we have identified new metabolic pathways implicated in IVIG resistance in KD. These pathways could be used as biomarkers to segregate IVIG nonresponders from responders prior to IVIG infusion. Also, boolean analysis of the transcriptomic data could be further extended to identify a universal biomarker that might predict IVIG response in other autoimmune diseases.