Tech Session I: Artificial Intelligence and Predictive Models
Physics Informed One-stop AI Preformulation Prediction Platform
Tuesday, July 15, 2025
11:49 AM – 12:00 PM EDT
Introduction: The surge of AI-driven drug discovery has flooded development pipelines with hit compounds. However, these increasingly complex molecules face significant developability challenges and formulation design demands. High-throughput and one-stop preformulation investigation tools are urgently needed to reduce development attrition. Although some preformulation properties have been mentioned in molecular machine learning, existing studies are limited by data quality, chemical space coverage, and often overlook drug solid-state characteristics and pharmaceutics-specific issues.
Learning Objectives:
At the completion of this activity, participants will know
Recognize the value of the physics informed TabPFN model in pharmaceutics scenarios.
Leverage batch preformulation and developability exploration with the one-stop AI platform.
Utilize interpretable formulation descriptors for AI-driven drug development.
