Finding and targeting cryptic pockets could dramatically expand the druggable proteome. However, discovering these sites remains challenging since they are only open a fraction of the time. It is also difficult to predict the functional relevance of a cryptic site as this often requires insight into allostery. Here we introduce attention enabled (AE-)PocketMiner, an artificial intelligence (AI) method that uses a graph neural network with an attention mechanism to simultaneously predict the locations of cryptic pockets and their allosteric coupling to the rest of the protein from a single input structure. We show that AE-PocketMiner outperforms past methods for identifying cryptic pockets and recapitulates known allosteric interactions. Moreover, we experimentally confirm newly predicted cryptic pockets and mutations that allosterically control pocket opening. AE-PocketMiner thus provides a powerful framework for multiple steps of the drug discovery process—including pocket identification, prioritization, and assay design—that will help expand the druggable proteome.
β-Lactamases confer bacterial resistance by hydrolyzing β-lactam antibiotics. Klebsiella pneumoniae carbapenemases (KPCs) are a major cause of resistance to carbapenems, but combinations of β-lactams with β-lactamase inhibitors, such as ceftazidime/avibactam, initially restored activity against these enzymes. However, treatment has driven the emergence of KPC variants that evade inhibition. Here we investigate the evolutionary and mechanistic consequences of three clinically observed substitutions -D179Y, T243M and H274Y- that together convert the broad-spectrum KPC-2 enzyme into a variant specialized for ceftazidime/avibactam resistance. Using steady-state and pre-steady-state kinetic analyses with ceftazidime, cephalothin and imipenem, we show that T243M enhances cephalosporin hydrolysis, whereas H274Y modulates this effect. Introduction of D179Y decreases both kcat and KM and shifts the rate-limiting step of catalysis to deacylation. Together, these mutations act cooperatively, reducing catalytic efficiency while promoting resistance to ceftazidime/avibactam. These findings reveal how epistatic interactions shape the evolution of clinically relevant β-lactamase variants and may help anticipate future resistance trajectories and guide inhibitor design.