TripK_a: Accurate and Scalable Acid-Base Dissociation Property Prediction via Triplet Interaction Networks and...

The acid-base dissociation constant (pK_a) characterizes a molecule's tendency to donate or accept protons, thereby fundamentally influencing its physicochemical properties and behavior. Consequently, pK_a is pivotal to diverse applications in drug discovery, including virtual screening and drug design. However, accurate pK_a prediction remains challenging, due to intricate atomic interactions among ionizable groups and the scarcity of experimental data. To address these, we propose TripK_a, a quantum-informed aqueous pK_a predictor that employs a triplet interaction network to model complex thermodynamic coupling in multisite protonation equilibria. TripK_a is fine-tuned using quantum-level descriptors as training targets, enabling the extraction of physical insights into molecular acid-base dissociation properties from limited high-quality pK_a data. Our proposed TripK_a outperforms state-of-the-art methods on multiple pK_a benchmarks, including the SAMPL6-8 and Novartis data sets, achieving reductions of over 10% in both mean absolute error and root-mean-square error. Beyond pK_a prediction, TripK_a further serves as a foundational model for diverse molecular properties prediction and interaction modeling tasks, benefiting from its chemically informed pK_a-pretraining.