RESEARCH ARTICLE Predicting response to mu l o

is available at the end of the articleBackground Patients suffering from cancer are usually treated with a range of chemotherapeutic agents, but the treatment efficacy varies greatly between patients. As new thera-peutic possibilities emerge, it is becoming increasingly important to identify individual patients who are un-likely to respond satisfactorily and who may benefit from carefully selected agents [1]. Resistance gene signatures (REGSs) for prediction of chemoresistance have been investigated extensively since the development of microarrays. The REGS can be grouped into classifiers and predictors where the classi-fiers assign a probability for each patient as sensitive or resistant, and the predictors assign each patient a numeric value where higher values indicate greater drug resistance. Studies generating REGS can either be per-formed by analysis of clinical data generated in vivo followed by a prognosis based reverse-translational ap-proach, or by analysis of laboratory data generate