Prostate cancer is the second leading cause of cancer death among men in the U.S., making it one of the most lethal malignancies. Predicting its course has been challenging, since only a fraction of patients experienced cancer recurrence after radical prostatectomy or radiation therapy. Researchers at the University of Pittsburgh have found a new way to better predict the outcomes of prostate cancer by developing a fusion gene-based machine learning tool. By examining the expression of 14 fusion genes in 607 prostate cancer samples, their results showed that fusion genes consistently improved the prediction rate of prostate cancer recurrence by Gleason score and serum PSA level, or the combination of both. These improvements occurred in both training and testing cohorts and were corroborated by multiple models.