Driver mutations can contribute to the initial and process of cancer.
Allostery regulates protein function at an allosteric site away from the functional site.
Thus, identifying the driver mutations in allosteric sites will be greatly benefit for
deciphering the mechanism of cancer and developing allosteric drugs.
Here, we provide a platform called DeepAlloDriver to predict the driver
mutations by a deep-learning method, which exhibited >93% performance in
both accuracy and precision. DeepAlloDriver will facilitate the revelation
of the underlying mechanism during cancer progression and helps to prioritize
cancer therapeutic targets.