Predicting Enzyme pH Optima from Structure Using Equivariant Graph Neural Networks

Abstract

Enzyme activity and stability are strongly modulated by pH, making the catalytic pH optimum (pH opt) a key parameter in enzyme development and biotechnological applications. Experimental determination of pH opt is, however, labor-intensive and time-consuming, motivating the development of accurate computational prediction methods. Here, we introduce pHoptNN, an E(n)-equivariant graph neural network designed to predict enzyme pH opt directly from three-dimensional protein structures. pHoptNN was trained on a curated dataset comprising nearly 12,000 enzymes with experimentally determined pH opt values and high-confidence structural models obtained from the Protein Data Bank and AlphaFold3. The model represents enzymes as atomic-level molecular graphs, integrating structural, chemical, and electrostatic features. Model development was guided by extensive hyperparameter optimization using genetic and Bayesian search strategies. On a held-out test set, pHoptNN achieved a root-mean-square error (RMSE) of 0.588 pH units, substantially outperforming the sequence-based method EpHod (RMSE = 0.879). Moreover, pHoptNN maintains robust predictive performance across different enzyme classes and pH ranges. These results demonstrate the utility of structure-based equivariant deep learning for enzyme pH opt prediction and highlight the potential of pHoptNN to accelerate enzyme discovery and engineering workflows.