借AI发力 打开“智造”新药大门

Core Viewpoint - The FDA is phasing out animal testing requirements for monoclonal antibodies and other drugs, shifting towards more efficient and human-relevant methods such as AI predictive models, organoids, and organ-on-a-chip technologies. This raises questions about the preparedness of the Chinese pharmaceutical industry for this transition [1]. Group 1: AI in Drug Development - AI can design molecular drugs and predict their clinical trial success rates and market potential, addressing the high failure rate of new drugs, which is around 90% during clinical trials [2]. - The National Cancer Center in China has established a dedicated institution to gather research data on rare diseases and new cancer targets, collaborating with AI companies and academic teams to enhance clinical research under AI guidance [2]. - AI technologies developed in recent years can provide optimal drug design solutions for multiple targets, significantly reducing trial and error costs in drug development [2]. Group 2: Enhancing Originality and Efficiency - Clinical trials account for over 70% of the time and costs in drug development, with traditional methods often lacking accuracy and reliability in efficacy predictions [4]. - Machine learning models have improved the predictive accuracy of drug clearance rates from a maximum of 65.8% to 94.1%, demonstrating the potential of AI to enhance safety assessments [4]. - AI can accurately estimate various parameters, including safety doses, thereby improving the capabilities of original drug development in China [4]. Group 3: Revitalizing Existing Drugs - AI can also uncover new functions of existing drugs or "rescue" failed drugs by identifying new therapeutic targets, leveraging the safety profile of older medications [5]. - Chinese research teams are continuously exploring new associations between drugs, diseases, and targets, providing innovative pathways to revitalize existing drug resources [5].