Revolutionsargues that new insights often come from scientific renegades who champion new paradigms to account for observations that cannot be adequately explained by existing theory1. His views intrigued me while I was a Princeton undergraduate interested in the history of science. Imagine the scientific upheaval in 1543 when Copernicus proposed a heliocentric model to explain various astronomical observations, challenging the geocentric view of Ptolemy that had been in place for centuries. Cancer biologists and physicians completing their training today are likely to assume that the imatinib story—the discovery that imatinib (Gleevec) is an effective treatment for chronic myeloid leukemia (CML)—is a logical extension of earlier landmark discoveries that CML is caused by BCR-ABL, the tyrosine kinase inhibited by imatinib. Although true in a broad sense, the backstory is not quite so simple. In 1995, the year that imatinib was first described by Nick Lydon and his colleagues, the general consensus was that cancers such as CML could be initiated by single oncogenic events or driver mutations. But there was skepticism about whether such tumors would remain dependent on the initial lesion, owing to the innumerable additional oncogenic events that accumulate in most cancers. This widely held view had important implications, because it predicted that inhibitors targeting the initiating lesion would fail unless a cocktail of drugs could be developed to target multiple lesions. Just 14 years later, this conclusion has been turned on its head. Most cancer drug discovery efforts are now focused on targeting individual oncogenic lesions, in the belief that many cancers remain dependent on driver mutations. Although the success of imatinib was not a revolution in the Copernican sense, it spawned a transformation in cancer research that has fueled an urgency to characterize cancer genomes comprehensively and discover the driver mutations in all cancers. My fascination with CML began during my clinical residency while I was caring for young patients undergoing allogeneic bone marrow transplantation (BMT). Although BMT offered a chance of cure, complications from graft-versus-host disease were substantial. BMT could be made safer by removal of T cells from the donor marrow, but the lack of graft-versushost disease was accompanied by a higher relapse rate, establishing a crucial role for the donor immune system in eliminating residual CML cells2. In parallel with these immunological insights into the mechanism of CML cure, several laboratories reported that the Philadelphia chromosome, originally linked to CML by Peter Nowell and his colleagues3 and characterized as a reciprocal translocation by Janet Rowley4, targeted the Abelson tyrosine kinase5, 6. In 1985, the year I began my internal medicine training, Owen Witte’s group at the University of California–Los Angeles (UCLA)