Writing in the July 12, 2021 online issue of Nature Communications, researchers at University of California San Diego School of Medicine describe a new approach that uses machine learning to hunt for disease targets and then predicts whether a drug is likely to receive FDA approval.