Southwest: AI empowers the future of pharmaceuticals across the entire process, with drug validation promising a bright future

Southwest released a research report stating that the AI pharmaceutical industry has gone through the embryonic and development stages, and since 2024 has entered a mature, deepened stage of commercialization validation. The core of competition has shifted from model algorithms to high-quality data, experimental closed-loop capabilities, and the ability to realize commercialization of AI drugs. AI pharmaceuticals are at a critical stage of commercial value validation, and it is recommended to capture investment opportunities from three major dimensions: R&D, commercialization, and BD. The bank recommends focusing on three main lines: technical barriers, implementation progress, and performance elasticity. Southwest's main points are as follows: How does AI empower the entire drug R&D process? AI applications have upgraded into a full-chain productivity tool covering "drug discovery - clinical development - commercialization - manufacturing supply," with core value reflected in cost reduction and efficiency improvement: the drug discovery stage can save 70%-90% of time, the preclinical and clinical stages can save 50%-80% and 50%-60% of time respectively, and cumulatively can reduce total R&D costs by about 50%. How does AI empower specific segments/areas of drug R&D? In specific segments, target discovery relies on knowledge graphs and multi-omics integration to break through "undruggable" bottlenecks; virtual screening uses machine learning acceleration to achieve second-level pre-screening of billion-scale compound libraries; molecular generation has entered the era of 3D diffusion models, directly generating adapted molecules in target space; ADMET prediction enables rapid early screening of druggability; DMTA automated closed loops connect virtual design with physical synthesis; on the clinical side, intelligent patient matching and adaptive trial design solve the pain points of difficult enrollment and high costs. At present, AI has covered the R&D optimization of all categories of drugs, including small molecules, antibodies, small nucleic acids, mRNA, and CAR-T. What are the core competitive barriers of AI pharmaceuticals? The industry has formed three core barriers: computing power, data, and models, with data and models more important than computing power. The dry-wet closed loop is the ultimate high-level barrier integrating the three major barriers. What is the investment logic of AI pharmaceuticals? AI pharmaceuticals are at a critical stage of commercial value validation, and it is recommended to capture investment opportunities from three major dimensions: R&D, commercialization, and BD. First, on the R&D side: pay attention to core technology breakthroughs and clinical data disclosure milestones, especially Phase II/III druggability validation, as well as data catalysts from industry academic conferences. Companies with dry-wet closed-loop capabilities have more prominent long-term barriers. Second, on the industry chain side: pay attention to upstream "pick-and-shovel" players that benefit from rising wet-lab demand, which are core beneficiaries during the AI pharmaceutical validation period. Third, on the BD side: BD transactions between global pharmaceutical companies and AI platforms continue to be active, and large-scale collaborations and License-out deals are important stock price catalysts. Related targets: Around the three main lines of technical barriers, implementation progress, and performance elasticity, it is recommended to focus on the following directions and related companies: 1) Full-stack platform leaders, prioritizing targets with outstanding dry-wet closed-loop capabilities: companies with end-to-end AI R&D capabilities and proprietary automated wet-lab systems have the deepest long-term barriers, INSILICO, XTALPI; 2) Beneficiaries of the AI industry chain, capturing certainty of performance realization: CXO companies that incorporate AI technology upgrades are expected to be the first to enjoy industry demand dividends, such as CRO+AI platforms Hitgen Inc., PharmaResources, Jiangsu Hualan New Pharmaceutical Material; comprehensive CXO companies WuXi AppTec, Pharmaron Beijing, CDMOs Asymchem Laboratories, WUXI BIO, Porton Pharma Solutions, clinical CRO Hangzhou Tigermed Consulting. At the same time, upstream sequencing, reagents and consumables, and other industry chain segments are expected to benefit first, including gene synthesis GENSCRIPT BIO, Shanghai Obio Technology, reagents and consumables NanJing Vazyme Biotech, etc., serum media Shanghai OPM Biosciences, etc., recombinant protein reagents Acrobiosystems, etc., model animals Biocytogen Pharmaceuticals, Shanghai Model Organisms Center, Inc., etc.; 3) Leaders in differentiated technology tracks, scarce targets in niche areas: focusing on the differentiated tracks of AI formulations and nano-delivery, Jitai Technology; 4) Local Pharma leaders that build their own AIDD platforms, with AI empowering internal new drug pipeline iteration: Jiangsu Hengrui Pharmaceuticals, SBP GROUP, Shanghai Fosun Pharmaceutical, CSPC PHARMA, etc. Risk warnings: risk of AI technology iteration falling short of expectations, risk of innovative drug R&D failure, risk of clinical progress falling short of expectations, risk of changes in industry regulatory policies, risk of fluctuations in computing power supply, etc.