Michael’s current research focus is to understand the etiology of human cancers. Although much is known about individual oncogenes and how they contribute to cancer, very little is understood about how these individual oncogenes cooperate to cause malignancy. To this end Michael is developing computational approaches to infer the combinations of oncogenic events that are required for malignancy. He has also worked on developing GRAPE, a computational method for detecting abnormal pathway behavior from gene expression profiles. GRAPE is a template-based method and is robust to batch effects.