Abstract
The integration of artificial intelligence (AI) in drug discovery promises to revolutionize the field by accelerating the development of effective and safe drug candidates. Despite increasing investment and partnerships in this space, the reality of implementing AI remains complex. In drug discovery chemistry, AI holds immense potential in generative chemistry, predictive modeling, and retrosynthetic analysis, yet challenges persist in ensuring the relevance, stability, and synthesizability of AI-derived structures. Demonstrating the value of AI requires robust, reproducible evidence from diverse projects, overcoming skepticism fueled by overhyped case studies. For widespread adoption, AI tools must be user-friendly, scalable, and capable of handling realistic data sets. Organizational changes are necessary to integrate AI into existing workflows effectively, augmenting the capabilities of chemists rather than replacing them. Overcoming these challenges and successfully deploying AI can significantly enhance decision-making in drug discovery, as evidenced by higher success rates in early clinical trials for AI-derived candidates.
Get full access to this article
View all access options for this article.