How machine learning revealed new targets for cancer drugs

"The bottom line is that we need new ways to analyze the massive amounts of cancer genome data, and this [method] is one that looks like it can lead to exciting and unexpected discoveries, and maybe treatments someday," said Dr. Bruce Clurman, one of the lead researchers on the work. The study was published Monday in the Proceedings of the National Academy of Sciences of the United States of America. Clurman, executive vice president and deputy director of Fred Hutchinson Cancer Research Center and holder of the Rosput Reynolds Endowed Chair at the Hutch, described the study and its implications in the following interview, which has been edited for brevity and clarity. This is a protein that degrades a lot of other proteins that drive cancer. It's called a tumor suppressor.