M$^2$DDI: A Unified Framework for Dynamic Multimodal Fusion in Drug-Drug Interaction Prediction

Runqing Xu, Siyi Liu, Yongqi Zhang

Keywords: AI for Science

Abstract: Drug-drug interaction (DDI) prediction is critical for ensuring patient safety and optimizing therapeutic outcomes. Existing computational approaches are limited by their inability to jointly model the heterogeneous mechanisms underlying DDIs, which span molecular structure, pharmacodynamic function, and network-mediated relations. To address this limitation, we introduce \texttt{M$^{2}$DDI}, a unified framework for dynamic multimodal fusion in DDI prediction. \texttt{M$^{2}$DDI} utilizes a Mixture-of-Experts architecture, with each expert dedicated to a distinct pharmacological modality. A novel prior-enhanced dual-path gating strategy adaptively selects relevant experts for each drug pair by integrating mechanism-matched feature queries and ATC-based biomedical priors, thereby aligning expert selection with underlying pharmacological mechanisms and addressing the challenge of data incompleteness. Empirical evaluation on benchmark datasets demonstrates that \texttt{M$^{2}$DDI} achieves state-of-the-art performance, particularly in new drug scenarios. Additional robustness experiments show that \texttt{M$^{2}$DDI} maintains high predictive accuracy even when modality-specific information is partially missing, outperforming existing methods under similar conditions. Analysis of expert selection patterns further confirms alignment with established pharmacological mechanisms. These results establish \texttt{M$^{2}$DDI} as an effective and mechanism-aware solution for comprehensive DDI prediction. The code are available at \url{https://anonymous.4open.science/r/M2DDI-AECB}

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 8457