Pangu Drug Molecule AI Model Accelerates New Drug Development as Global Tech Giants Race to Deploy 'AI + Healthcare'
-
The innovative integration of AI and biopharmaceuticals has long been highly anticipated. With the support of tech companies like Huawei, advanced AI technologies have become a key driver for high-quality development in the biopharmaceutical industry across Shenzhen and the Greater Bay Area.
Pangu Drug Molecule AI Model Accelerates New Drug Development
The pharmaceutical industry follows the "Rule of Tens"—it typically takes over 10 years and $1 billion to develop a new drug, with only about 10% of candidates reaching clinical approval. Now, with AI models, improving the efficiency and quality of drug development is no longer an "insurmountable challenge."
Professor Liu Bing from Xi’an Jiaotong University’s First Affiliated Hospital experienced this firsthand. His team discovered a broad-spectrum antibiotic by leveraging Huawei Cloud’s Pangu drug molecule model for lead compound optimization. Using Huawei’s medical AI platform, they screened 300 million small-molecule drugs in under a week, identifying the first new antibiotic target in 40 years. This breakthrough cut lead compound development time to one month and reduced costs by 70%.
Similarly, Shenzhen-based Microcyto Pharma used Huawei’s Pangu model and AI platform to boost tumor drug design efficiency by one-third, with molecular binding energy improving by 40%, significantly accelerating oncology research.
Launched in 2021, Huawei Cloud’s Pangu model was then the industry’s largest drug molecule AI, offering capabilities like binding prediction, property forecasting, and molecular optimization. In July 2023, its 3.0 upgrade further enhanced key technical features for biopharmaceutical applications.
The model was trained on 1.7 billion drug molecules, achieving 20% higher accuracy in druggability prediction than traditional methods. It also generated 100 million novel drug-like molecules, with 99.68% structural uniqueness, expanding possibilities for new drug discovery.
Global Tech Giants Race to Deploy 'AI + Healthcare'
AI’s potential in healthcare spans accurate diagnostics, personalized treatment, and intelligent assistance. Giants like Google, Amazon, NVIDIA, Tencent, and JD.com are aggressively investing in this space.
While generative AI is transforming medicine, challenges like proprietary algorithms, data privacy, and diagnostic precision remain. Previously focused on medical imaging, AI’s deeper integration with clinical data now promises broader breakthroughs.
NVIDIA, for instance, invested $50 million in biotech firm Recursion to advance AI models in chemistry and biology, boosting Recursion’s stock by 78% and its own by 3.5%.
'AI + Healthcare' Enhances Clinical Efficiency
The 'Cloud Ruijin' smart service platform features a fully managed intelligent operation system. Its multi-campus smart imaging cloud platform, empowered by large models, enables AI-assisted diagnosis across multiple radiology and pathology specialties, enhancing clinical efficiency.
The 'China AI Large Model Map Research Report' reveals that China has released 79 large models with over 1 billion parameters, with applications expanding from office, lifestyle, and entertainment to healthcare, industry, and education. At the 'Boundless Love, Daily Renewal' AI forum in Shanghai, Professor Dimitris Metaxas from Rutgers University noted that medical large models excel in multimodal data analysis and addressing clinical long-tail problems, significantly improving AI performance in data-scarce scenarios.
For healthcare services, SenseTime's 'Great Doctor' medical language model powers a smart healthcare platform with features like intelligent triage, pre-consultation, pre-examination, follow-up, and hospital navigation, digitizing and streamlining the patient journey.
SenseTime's SenseCore boasts 5000 Petaflops of computing power, supporting simultaneous training of 20 ultra-large models with 100 billion parameters. This enables efficient model training for clinical long-tail needs and assists medical institutions in autonomous model development.
AI has revolutionized healthcare in recent years, offering immense potential. McKinsey estimates AI could generate $3.5–5.8 trillion annually, with the global AI healthcare market projected to reach $127 billion by 2025, accounting for 20% of the total AI market.
ReportLinker predicts the global healthcare AI market will grow from $14.6 billion in 2023 to $102.7 billion by 2028, a 47.6% CAGR. In China, AI healthcare has grown 40–60% annually since 2019, with core software nearing ¥3 billion and robotics pushing the total to ¥6 billion.
Zhongtai Securities highlights the vast potential of AI in healthcare/pharma, particularly in AI pathology, imaging, and drug discovery.