A molecular model of the big-conductance potassium (BK) channels is of great interest as BK channels are a crucial member of the cellular communication network. It is implicated in numerous diseases including muscular dystrophy, epilepsy and stroke, yet remains a challenging drug target since the details of the gating mechanism remain unclear. But importantly, there are cryo-EM structures of both the open and closed conformations, and over 470 mutations have been functionally annotated. Leveraging these structures, I constructed a physics-based description of the effect of each possible mutation using MD simulations and computational mutagenesis, then trained these descriptors on the mutagenesis data set using machine learning methods. The result is a quantitative model predicting the function of any BK mutation, which shows good agreement with known trends including hydrophobic dewetting. We further validated our predictions by testing four mutant channels in the laboratory of our collaborator Jianmin Cui at Washington University in St. Louis. We found remarkable agreement among the four mutants we tested, which were at a particular site on the S5-S6 interface, and we now strongly believe this site to be crucial for mediating information between the VSD and PGD and will likely feature in future investigations.(as of Aug 2023) The paper has been accepted to PLOS Comput Biol. You can check out the preprint here or the code on GitHub.