Computational insights into drug hygroscopicity by coupling machine learning and molecular simulation.

of 0.701 ± 0.075 for regression of moisture-induced weight change (%), and accuracies of 0.741 ± 0.047 and 0.872 ± 0.029 for four-class and binary classification. SHapley Additive exPlanations (SHAP) analysis identified molecular surface area, polarity, and electrostatic descriptors as key factors influencing hygroscopicity. Building upon this insight, molecular dynamics and quantum chemical simulations further revealed that polar functional groups, hydrogen bonding, and surface conformations govern water-molecule interactions, consistent with the ML-derived insights. Overall, the effective combination of AI-driven predictions and physics-based mechanistic insights underscores the potential of this approach for preformulation developability screening and optimization, offering a promising avenue to reduce R&D costs and enhance drug development efficiency.