AI-Powered Drug Discovery Accelerates in Pudong: XtalPi and Insilico Medicine Advance Innovative Therapeutics

(Source: Hello Zhangjiang)

The innovative development of the biopharmaceutical industry relies heavily on the efficient empowerment of artificial intelligence technologies and innovative tools. XtalPi recently announced that PEP08, a next-generation PRMT5 inhibitor discovered with its support in collaboration with Zhiheng Biotech, has reached a significant clinical research milestone. PEP08 has been approved for Phase I clinical trials in solid tumors by the Australian Human Research Ethics Committee, the Australian Therapeutic Goods Administration (TGA), and Taiwan's "Food and Drug Administration" under the Ministry of Health and Welfare.

In recent years, artificial intelligence technology has been widely applied to various stages of pharmaceuticals, improving the efficiency and quality of new drug development while significantly reducing R&D costs. The market size for AI-driven pharmaceuticals continues to expand. Reporters have learned that many companies in Pudong are increasingly focusing on "AI+", leading to a surge in innovative drug products, and the industry is welcoming broader development opportunities.

AI+RobotDrug R&D Platform Empowers Drug Innovation Efficiently

PRMT5 (Protein Arginine Methyltransferase 5) is a key enzyme that is overexpressed in various cancers and promotes tumor growth. In tumors with homozygous deletion of MTAP (methylthioadenosine phosphorylase), inhibiting PRMT5 activity can produce a "synthetic lethality" effect, offering the potential for highly efficient and safe targeted cancer therapy. However, traditional PRMT5 inhibitors often cause toxicity issues due to poor selectivity, representing a significant unmet clinical need.

After the project launch, XtalPi, with its intelligent drug discovery platform that integrates physical models, AI algorithms, and robotic experiments, combined with the expert knowledge of Zhiqing Biotech Pharma, quickly initiated rational drug design. XtalPi generated a molecular library of millions for this project, from which it accurately screened a group of lead compounds with excellent drug activity and selectivity, as well as novel molecular scaffolds. After multiple rounds of ADMET property evaluation and iterative optimization combining physical and AI models, PEP08 was ultimately selected for preclinical development.Candidate compounds were completed in preclinical development by Engine Biosciences.

Preclinical study data shows that, compared with first-generation non-selective PRMT5 inhibitors, PEP08 demonstrates significant advantages in toxicity and safety, possesses excellent blood-brain barrier penetration ability, and exhibits ideal overall drug-like properties. In multiple animal efficacy models, PEP08 has demonstrated potent in vivo efficacy. Compared with other second-generation PRMT5 inhibitors currently in clinical development, the comprehensive pharmaceutical properties of PEP08 show potential best-in-class effects and broad prospects.

PEP08 Successfully Obtains Clinical Trial Approval from Regulatory Authorities in Two Regions and Achieves Milestone Payment, Marking an Important Phase Achievement in the Collaboration between XtalPi and Pharmanest Biopharmaceuticals. This Once Again Demonstrates the Capability of XtalPi's Platform to Efficiently Empower Drug Innovation.

A relevant responsible person from XtalPi stated that they will continue to rely on the AI + robotic drug discovery platform to help more partners accelerate the discovery of innovative drugs with market potential and differentiated competitiveness, bringing more patients high-quality and efficient treatment options.

Driving the First Conceptual Validation of AI-Driven Drug Discovery

AI is rapidly transforming the pharmaceutical industry, reshaping the landscape from target discovery to precision medicine and bringing unprecedented opportunities to accelerate drug development and market entry. Cutting-edge companies in Pudong have significantly increased the proportion of AI-driven new drug development projects through self-developed platforms or collaborative innovation models.

For instance, Insilico Medicine, a "pioneer" in the AI pharmaceutical field, generated 3.5 billion yuan in revenue in 2024 at its Shanghai company located in Pudong, marking a 108% year-on-year increase. Insilico Medicine is a clinical-stage biopharmaceutical company driven by generative artificial intelligence. In January 2025, Insilico Medicine announced the establishment of its China Technology Headquarters in Pudong, collaborating with upstream and downstream enterprises to create a full-cycle life science and health management closed-loop ecosystem platform in Pudong that integrates bioinformatics collection and analysis, pharmaceutical R&D and production, and clinical diagnosis and treatment. Alex Zhavoronkov, founder and CEO of Insilico Medicine, stated that the company is planning and building a fully automated laboratory and its China Innovation Headquarters in Pudong. "Looking ahead, the support from the AI platform will not be limited to biopharmaceuticals but will also extend to several promising frontier fields such as contrast agents, new materials, new energy, and anti-aging."

On June 3, the first clinical proof-of-concept results of AI-driven drug discovery in the industry were published in *Nature Medicine*. Insilico Medicine and its co-authors jointly released the safety and efficacy data from the Phase IIa clinical trial of the TNIK inhibitor Rentosertib (ISM001-055).

Rentosertib is a potential global first-in-class small molecule drug for the treatment of idiopathic pulmonary fibrosis (IPF), discovered and designed through the Pharma.AI generative AI platform empowered by Insilico Medicine. Exploratory biomarker analysis in the study further validated the biological mechanism of the novel target TNIK identified via artificial intelligence methods, with results supporting Rentosertib's potential anti-fibrotic and anti-inflammatory effects. The target identification and molecular design of Rentosertib were both driven by artificial intelligence, representing true innovation in the pharmaceutical industry.

Previously, Insilico Medicine has announced several key milestones, demonstrating that its self-developed AI platform can significantly reduce costs and enhance the efficiency of early-stage drug discovery. Through the deep integration of AI and automation technologies, the company has established over 30 self-developed pipelines. As multiple AI-driven R&D pipelines progress in parallel, Insilico Medicine is also poised to become one of the first companies to successfully advance and complete clinical validation for AI-discovered and AI-designed drug candidates.

Author: Yang Zhenying