Cancer Driver Mutations Predicted With Allosteric Site-Informed Deep Learning Approach

For a paper appearing in Nucleic Acids Research, researchers at the Shanghai Jiao Tong University School of Medicine and other centers in China describe a cancer driver prediction method called DeepAlloDriver that relies on deep learning for cancer driver mutation detection aimed at allosteric drug target design. Starting with gene symbol, amino acid substation, and/or DNA mutation data, the team explains, DeepAlloDriver identifies allosteric sites where cancer-related alterations are expected to alter protein regulation, providing a cancer driver prediction score that takes protein structure into account to assess affected genes, proteins, and pathways, while flagging possible therapeutic approaches. In their paper, for example, the authors applied this approach to the NTRK1 and RRAS2 genes, identifying apparent allosteric driver mutations, suspected oncogenic mechanisms, and possible drug targets. "Allostery is a regulatory approach that transmits information in biological systems and can be utilized to decipher molecular mechanisms in a wide spectrum of biological processes and discover cancer driver proteins," they explain. "DeepAlloDriver provides an efficient service to help clinicians and biologists with better decision-making in identifying allosteric driver mutations and carcinoma-relevant targets."