生命科学家钱玥：AI是科研人员的强大助力，而非替代者

Core Insights - The lecture series "AI and the Future of Universities" at Fudan University highlighted the role of AI as an efficient auxiliary tool in accelerating drug development processes and reducing costs, rather than replacing researchers [3][7] - AI's application in drug design has evolved from basic machine learning concepts to breakthroughs in deep learning networks and the latest generation of large language models, which enhance the value extraction from vast data [3][7] Drug Design and AI Integration - The core of AI-driven drug design lies in the integration of "wet and dry experiments" and deep interdisciplinary collaboration, combining traditional fields such as structural biology, medicinal chemistry, and pharmacology with AI tools for feature extraction, prediction, and generation [4][8] - The wet lab processes include protein structure analysis, high-throughput screening, drug synthesis, and efficacy studies, providing essential biological data for AI models, which analyze and iterate based on physical principles and large datasets, creating a closed-loop system for exploring new drug targets and mechanisms [4][8] Efficiency and Success Rates - In the field of macromolecule drug development, AI significantly enhances efficiency; traditional antibody development requires lengthy processes involving animal immunization and humanization, while AI can directly generate candidate sequences, reducing the screening volume by 10 to the power of 7 to 8, with a success rate exceeding 85% and a 400% increase in efficiency [4][8] - AI serves as a powerful support tool for researchers in coding, literature review, and knowledge transfer, while human researchers' core values remain in interdisciplinary collaboration, scientific intuition, and innovative thinking [4][8]