PhD | Digital Drug Hunter | 8 years in industrial & academic Computer-Aided Drug Design (CADD) | Computational Chemistry | Generative AI for Drug Discovery
Having spent two years working in China's burgeoning life sciences sector (2017-2019), I've maintained a keen interest in its development. A question I often encounter is about the current state of computer-aided drug design (CADD) and artificial intelligence in the country's pharmaceutical landscape. Public information can be fragmented, so I wanted to offer an overview of a sector that is rapidly becoming a global force.
A market ignited: growth and global megadeals
China's AI-driven drug discovery sector is not just growing; it's exploding. While the global market hit $1.76 billion in 2024 with a compound annual growth rate (CAGR) of over 30%, China's market is expanding at an even more astonishing rate. According to ChinaDaily, it grew from a modest $9.74 million in 2019 to over $410 million in 2023, marking a CAGR of 57%. Projections suggest this blistering pace will continue, with the market potentially exceeding 5.86 billion yuan by 2028. According to Grand View Research, whole China's biotechnology market is projected to grow from approximately $74 billion in 2023 to nearly $263 billion by 2030, reflecting a compound annual growth rate (CAGR) of 19.8% from 2024 to 2030.
This rapid growth is underscored by a series of landmark collaborations that signal the world's confidence in Chinese innovation. In 2025 alone, the industry has seen several multi-billion-dollar deals:
Syneron Bio inked a deal worth up to $3.4 billion with AstraZeneca to develop macrocyclic peptides for chronic diseases.
Helixon Therapeutics, through its U.S.-affiliated entity Earendil Labs, entered a partnership with Sanofivalued at over $1.8 billion to develop bispecific antibodies for autoimmune diseases.
In one of the largest license-out deals in China's biopharma history, XtalPiannounced a partnership with U.S. biotech DoveTree, potentially worth up to $6 billion. This collaboration will deploy XtalPi's AI and robotics platforms for preclinical research, with the goal of delivering ready-to-test clinical candidates directly to DoveTree's labs.
These agreements demonstrate that Chinese AI drug-discovery capabilities are gaining significant global recognition.
From CADD 1.0 to the AI "3.0 Era"
The engine driving China's biopharma ascent is a profound technological paradigm shift. The landscape is rapidly evolving from the established methods of traditional computer-aided drug design (CADD) to a new era defined by the creative power of AI. This progression can be understood in distinct phases.
As Ren Feng, co-CEO of Insilico Medicine, has articulated, the industry has moved from a "CADD 1.0" phase, dominated by physics-based simulations, into the current "AI 2.0" era. This new phase, which began gaining momentum around 2013, is characterized by wide adoption of generative AI. The crucial difference is the shift from pure analysis to creation; algorithms can now design novel molecules from scratch (=de novo design), rather than simply screening or refining existing ones.
Now, the industry is on the cusp of the "3.0 era": a future of end-to-end integration where AI is woven into the entire fabric of drug research and development. This vision extends from initial target discovery and validation, through molecular design and synthesis planning, all the way to intelligent clinical trial design and patient management. Chinese biotechs and "techbio" firms are not just participants in this evolution; they are at its forefront, aggressively adopting and innovating upon these advanced methodologies.
The foundational bedrock: advanced CADD techniques
Structure-Based and Ligand-Based Drug Design (SBDD/LBDD) - these are the cornerstones of modern rational drug design. In SBDD, scientists use the known 3D structure of a biological target (like a protein's "lock") to computationally design a drug molecule (the "key") that will fit perfectly, modulating its function. Companies like Viva Biotech have built deep expertise in this area.
When a target's 3D structure is unknown or of unsatisfactory quality, the strategy shifts to LBDD. This approach relies on the knowledge of existing molecules (ligands) that are known to be active to deduce the characteristics required for biological activity. This is accomplished through a variety of techniques. Some methods compare molecules based on their 3D shape similarity and the distribution of their electrostatic charges, operating on the principle that compounds with similar physical and electronic profiles are likely to bind to the same target. Other, more abstract methods involve building a pharmacophore model. This model acts as a 3D blueprint, defining the essential arrangement of chemical features, such as H-bond donors and acceptors, aromatic rings, and charged groups, that are critical for a molecule to function. Whether based on direct shape comparison or an abstract pharmacophore, these models then serve as powerful templates to guide the computational search for new and potentially more effective molecules.
Failing faster and cheaper: the power of in silico prediction
In Silico ADMET Prediction: One of the primary causes of late-stage drug failure is unforeseen toxicity or poor pharmacokinetics. To combat this, Chinese firms are heavily leveraging computational tools to predict a drug candidate's Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) profile. By running these simulations early, AI models trained on vast historical datasets can flag compounds with a high probability of failure long before they enter costly lab experiments or animal testing, embodying the crucial industry mantra of "fail fast, fail cheap."
AI-driven drug discovery (AIDD) platforms in China
This is where the most significant and disruptive innovation is occurring. Rather than being a single tool, AIDD platforms are integrated systems that automate and accelerate multiple stages of the discovery pipeline. This is the domain of leaders like XtalPi, Insilico Medicine, and CarbonSilicon. Their platforms are capable of:
Intelligent target identification: moving beyond known biology, these AI systems can analyze immense multi-omics datasets (genomics, proteomics, clinical data) to identify novel biological targets that may be the root cause of a disease, including those previously considered "undruggable."
De novo molecular design: This is the heart of generative AI. Instead of searching through virtual libraries of existing compounds, these algorithms perform a kind of computational alchemy, generating millions of entirely new, virtual molecules that are optimized from their inception for desired properties like high potency, low toxicity, and good selectivity.
High-throughput virtual screening and optimization: For a given target, AI platforms can screen virtual libraries of billions of compounds in a fraction of the time and cost of physical screening. They then rapidly iterate on the most promising "hits," computationally optimizing them into viable "lead" compounds with superior drug-like properties.
Synthesis prediction: A brilliant molecule on a computer is useless if it cannot be made in a lab. Advanced AI platforms can now also predict the most efficient chemical synthesis pathways (though it is still very challenging!, providing a practical roadmap for medicinal chemists and dramatically shortening the timeline from digital design to physical compound.
By integrating these capabilities, China's leading techbio companies are building a new, more rational, and vastly more efficient paradigm for discovering the medicines of tomorrow.
The engines of growth: policy, investment, and talent
Several key factors are fueling this rapid ascent of the Techbio sector in China:
Tangible government support: the Chinese government has made innovation in healthcare a priority. A blueprint for AI in healthcare issued in November 2024 specifically named intelligent drug discovery and AI-assisted clinical trials as key areas. National strategies like the "Healthy China 2030" initiative and the 14th Five-Year Plan are creating powerful policy tailwinds.
Massive investments: Both domestic and international capital are flowing into China's AI and biotech sectors, funding a boom in R&D.
A thriving tech ecosystem: China's robust technology infrastructure, including national supercomputing centers in cities like Jinan and Tianjin, provides the immense computational power required for modern CADD and AI.
The "sea turtle" effect: A significant number of Chinese scientists and entrepreneurs who have studied and worked abroad (often called "sea turtles") are returning, bringing with them advanced expertise and global perspectives that are crucial for innovation.
Hubs of innovation and the vanguard of "techbio"
China's CADD and AI drug discovery revolution is not happening uniformly; it is powerfully concentrated in dynamic, interconnected biotech hubs where government policy, investment, academia, and industry converge. These clusters create a "rainforest-style" ecosystem where innovation thrives.
Geographic clusters driving innovation
Three major geographic regions stand out as the nerve centers of this transformation:
The Yangtze River Delta (YRD) is a Chinese biopharma core. Encompassing Shanghai and the provinces of Jiangsu and Zhejiang, the YRD is China's most advanced and commercially dominant biopharmaceutical cluster.
Shanghai: at the heart of the YRD, Shanghai is a global hub for AI-based drug discovery, cell and gene therapies, and medical robotics. The city's Zhangjiang Hi-Tech Park, often called "Pharma Valley," is a prime example of this concentration. It hosts over 1,400 biomedical companies, including the R&D headquarters for many of the world's top pharmaceutical giants like Roche, Pfizer, and Eli Lilly. The park is also home to an "AI Island" and an AI pharmaceutical R&D alliance, explicitly designed to foster collaboration and accelerate innovation. In a significant move, AI pioneer Insilico Medicine announced plans in early 2025 to establish its China technology headquarters in Shanghai's Pudong New Area, aiming to create a full-cycle R&D platform.
Suzhou and Hangzhou: these neighboring cities are critical components of the YRD ecosystem. Suzhou, a major hub for biotech events like BIOCHINA, is rapidly growing its biopharma industry, supported by provincial plans to accelerate innovation across the entire industrial chain.
A bird's eye view over the AIsland located in the Zhangjiang Science City in Shanghai.
The Greater Bay Area (GBA) is where Tech meets Biotech. Integrating nine cities in Guangdong province with Hong Kong and Macao, the GBA leverages its world-renowned strength in technology to fuel a new generation of life science companies.
Shenzhen: as China's original high-tech special economic zone, Shenzhen is now a key driver of the GBA's evolution into a global sci-tech innovation hub. The city is the headquarters for AI drug discovery leader XtalPi and the genomics powerhouse BGI. The government is actively supporting the creation of major innovation platforms in AI, life information, and biomedical laboratories in Shenzhen. The city's proximity to Hong Kong's financial markets and basic research talent further enhances its appeal.
Guangzhou and Hong Kong: Guangzhou is home to emerging biotechs, while Hong Kong provides a crucial link to global capital and foundational scientific research, creating a synergistic relationship that propels the entire region forward.
Beijing is the research and policy capital. As the nation's capital, Beijing combines top-tier academic and research institutions with significant policy-driven initiatives. The city is home to the Zhongguancun Life Science Park, which hosts over 500 medical and health enterprises, including notable companies like BeiGene. In a major recent development, Beijing unveiled plans for a new international pharmaceutical innovation park in its southern suburbs, specifically designed to attract global scientists and promote the integration of AI with new medicine R&D. This initiative, located in the Beijing E-Town, aims to become a global hub for medical innovation and a demonstration zone for AI applications by 2030.
The leading players - a new generation of "techbio"
While established CROs and contract development and manufacturing organizations (CDMOs) like WuXi Biologics are increasingly integrating AI and CADD into their service platforms, a new breed of company is leading the charge. These are the "techbio" firms - fundamentally technology-driven companies, building scalable AI platforms to make drug discovery more efficient and predictable.
XtalPi: a leader in integrating AI with quantum physics and robotics, XtalPi has a significant presence across China's major hubs, with its headquarters in Shenzhen and key research centers and automation labs in Shanghai and Beijing. The company is known for its high-profile partnerships with global giants like Pfizer, Eli Lilly, and its recent multi-billion dollar deal with DoveTree.
Insilico Medicine: a pioneer in using generative AI for drug discovery, Insilico has made China a cornerstone of its R&D strategy. While headquartered in the U.S., the company's founder has emphasized the importance of its operations in China for synthesis, testing, and clinical trials. Its new technology headquarters in Shanghai and labs in Suzhou underscore its deep commitment to the region. Insilico has successfully advanced multiple AI-discovered drug candidates into clinical trials in China, demonstrating the tangible results of its platform.
CarbonSilicon, BioMap, and iCarbonX: these companies represent the broader trend of AI-driven life sciences. CarbonSilicon provides a one-stop SaaS platform for drug discovery. BioMap focuses on using AI for real-time analytics. iCarbonX, based in Shenzhen, is a prominent player in leveraging AI and multi-omics data for precision medicine, backed by major investors like Tencent.
These hubs and the companies within them form a deeply interconnected network. The strategic co-location of top-tier academic institutions, innovative techbio startups, global pharmaceutical R&D centers, and state-of-the-art manufacturing facilities is a deliberate strategy, creating a fertile ground where the future of medicine is being coded.
Challenges on the horizon
Despite the impressive progress, the path forward is not without its obstacles. The industry faces several persistent challenges:
Computational and data hurdles: the accuracy of computational models and access to high-quality, standardized data remain critical areas for improvement.
The talent gap: there is a need for more specialists who possess expertise at the intersection of computational science and drug discovery.
An evolving regulatory landscape: navigating the regulations for AI-discovered drugs is a global challenge, as authorities work to create clear pathways for approval.
The final hurdle - commercialization: It is crucial to remember that while investment and preclinical activity are surging, no AI-discovered drug has yet received global approval. The commercialization loop remains incomplete, and the industry is still working to prove it can deliver on its ultimate promise.
Even with these challenges, the trajectory is clear. Through a powerful combination of government support, massive investment, a world-class tech ecosystem, and a new wave of innovative companies, China is rapidly cementing its position as an indispensable player in the future of global drug discovery. The next decade will be pivotal as these AI-driven candidates move through clinical trials and we see whether this technological revolution translates into a new generation of medicines.
References
Ting-ting Liu, Dong Lu, Hao Zhang, M. Zheng, Huaiyu Yang, and 5 more (2016). Applying high-performance computing in drug discovery and molecular simulation. National Science Review
Yumeng Lin, Youquan Zhang, Dongyang Wang, Bowen Yang, Ying-Qiang Shen (2022). Computer especially AI-assisted drug virtual screening and design in traditional Chinese medicine. Phytomedicine
H. C. S. Chan, Hanbin Shan, T. Dahoun, H. Vogel, Shuguang Yuan (2019). Advancing Drug Discovery via Artificial Intelligence: (Trends in Pharmacological Sciences 40, 592-604; 2019). TIPS - Trends in Pharmacological Sciences
Benjamin Siddiqui, Chandra Shekhar Yadav, Mohd Akil, Mohd Faiyyaz, Abdul Rahman Khan, and 6 more (2025). Artificial Intelligence in Computer-Aided Drug Design (CADD) Tools for the Finding of Potent Biologically Active Small Molecules: Traditional to Modern Approach. Combinatorial chemistry & high throughput screening
W. Warr (2017). A CADD-alog of strategies in pharma. Journal of Computer-Aided Molecular Design
Donglin Song (2025). Integration and Innovation of AIDD and CADD: Cutting-Edge Technologies and Future Trends in Accelerating Drug Development. Journal of Diseases and Medicinal Plants
Ahmad F Eweas, I. Maghrabi, Ali Ibrahim Namarneh (2014). Advances in molecular modeling and docking as a tool for modern drug discovery
Li Hong (2005). Virtual screening and new drug discovery
Yuxi Wang, Zelin Hu, Junbiao Chang, Bin Yu (2025). Thinking on the Use of Artificial Intelligence in Drug Discovery. Journal of Medicinal Chemistry
Ban Chen, Shuangshuang Liu, Huiyin Xia, Xican Li, Ying-Qing Zhang (2025). Computer-Aided Drug Design in Research on Chinese Materia Medica: Methods, Applications, Advantages, and Challenges. Pharmaceutics
Dongsheng Yang, Jian Li, W. Mak, Aole Zheng, Xiao Zhu, and 3 more (2025). PBPK Modeling: Empowering Drug Development and Precision Dosing in China. CPT: Pharmacometrics & Systems Pharmacology
Sun Hongmao (2015). A Practical Guide to Rational Drug Design
Chloe Markey, Samuel Croset, Olivia Ruth Woolley, Can Martin Buldun, Christian Koch, and 2 more (2024). Characterizing emerging companies in computational drug development. Nature Computational Science
Feng-Xu Wu, L. Zhuo, Fan Wang, Wei Huang, Gefei Hao, and 1 more (2020). Auto In Silico Ligand Directing Evolution to Facilitate the Rapid and Efficient Discovery of Drug Lead. iScience
Shaoliang Peng (2016). High performance computational biology and drug design on TianHe Supercomputers. IEEE International Conference on Bioinformatics and Biomedicine
Xiaoqin Tan, Jiacheng Xiong, Tingfei Zhu, Ziming Huang, X. Tong, and 11 more (2019). Development of drug design in China: 40 years of achievements
A. Sampene, Fatuma Nyirenda (2024). Evaluating the effect of artificial intelligence on pharmaceutical product and drug discovery in China. Future Journal of Pharmaceutical Sciences
Yuzhu Wang, Jia Ji, Ye Yao, Jing Nie, Fengbo Xie, and 2 more (2024). Current status and challenges of model-informed drug discovery and development in China. Advanced Drug Delivery Reviews
M. Guedj, Jack Swindle, Antoine Hamon, Sandra Hubert, Emiko Desvaux, and 9 more (2022). Industrializing AI-powered drug discovery: lessons learned from the Patrimony computing platform. Expert Opinion on Drug Discovery
T. D. A. Kumar, N. Swathi (2022). Machine learning empowered drug discovery. German Journal of Pharmaceuticals and Biomaterials
T. Stouch (2017). Urgency and austerity as drivers of success. Journal of Computer-Aided Molecular Design
Leming Shi, Z. Su, A. Xie, Chenzhong Liao, Wei Qiao, and 7 more (2006). An Integrated Biochemoinformatics System for Drug Discovery
I. Ghosh (2006). Target based high throughput screening and lead designing in pharmaceutical drug industry
Xiaoming Zhu (2018). Case I: 3DMed: An R&D Platform for Personalised Precision Anti-cancer Drugs
