The 146th Annual Meeting of the Pharmaceutical Society of Japan (Osaka)

Organizer: Hidetoshi Arima (Sch. Pharm., Daiichi Univ. Pharm.), Masakazu Sekijima (Dept. Comp. Sci., Science Tokyo)

In the rapidly evolving landscape of pharmaceutical sciences, the integration of heterogeneous data has become a critical challenge. Modern research in drug discovery, drug delivery systems (DDS), and medical diagnostics requires the simultaneous processing of diverse modalities, including chemical structures, 3D conformational data, physicochemical properties, multi-omics profiles, and clinical time-series data. Conventional single-modality analyses often fail to capture the complex, non-linear correlations inherent in biological systems. Consequently, multimodal AI has emerged as a transformative approach to synthesize these disparate data streams for holistic understanding. This symposium highlights the frontier of pharmaceutical research pioneered by multimodal data integration and AI/ML technologies. Prof. Masakazu Sekijima will present novel strategies for interaction-oriented screening and de novo molecular generation, utilizing energy decomposition analysis of protein-ligand interactions coupled with generative models. Prof. Yoshihiro Yamanishi will discuss a comprehensive AI framework that fuses disease-related omics, protein structural data, and chemical space to accelerate target identification and rational drug design. Prof. Hidetoshi Arima will introduce a data-driven approach to rational DDS design, applying AI to decipher hierarchical data from molecular recognition to cellular response within cyclodextrin-based supramolecular and conjugate systems. Finally, Prof. Manabu Kano will demonstrate the clinical utility of machine learning in analyzing physiological time-series data, such as glucose levels and electrocardiograms derived from wearable devices. Through these presentations, we aim to discuss how multimodal AI is redefining the methodologies of drug discovery and the future of medicine.