A Hangzhou-based biotechnology startup named MindRank has advanced to Phase 3 clinical trials with its obesity medication, becoming the first AI-enhanced Category 1 new drug in China to achieve this milestone.
Last month, MindRank revealed that it has started a Phase 3 study in China for MDR-001, an AI-assisted small-molecule GLP-1 receptor agonist. These types of drugs imitate natural hormones to control glucose levels and hunger.
As stated by Niu Zhangming, the founder and CEO of MindRank, the medication marked the initial AI-enhanced, Class I novel drug to attain such an advanced phase within China. The firm targeted regulatory clearance during the latter part of 2028, setting the stage for a commercial release in 2029.
Are you curious about the most significant issues and developments happening globally? Find your answers here with SCMP Knowledge Our latest platform featuring carefully selected content including explanations, frequently asked questions, analysis, and visual graphics, presented by our acclaimed team.
The newly developed medication took approximately four and a half years to create, whereas traditionally it would take between seven and ten years to achieve this phase, as stated by Niu. He further mentioned that improvements in efficiency brought about by artificial intelligence have significantly reduced total research and development expenses by at least 60 percent.
As per Niu, human experts can identify a specific target, usually a protein inside the body associated with a particular illness, and use their advanced AI technologies to quickly create possible medications. This process enables scientists to narrow down and choose the best options from those generated by artificial intelligence.
" It's similar to managing an automatic production line," Niu mentioned.
MindRank created a dedicated biomedical system using open-source large language models (LLMs) combined with Retrieval-Augmented Generation (RAG) techniques. RAG enables an LLM to access current data from outside sources.
This enhanced the precision of target research, increasing it from approximately 85 percent—the industry standard—to more than 97 percent, as reported by Niu. Target research accuracy refers to how well potential treatment targets for a condition are recognized.
According to Niu, AI can evaluate a medication's safety and effectiveness using sophisticated prediction models, handling intricate computations that were once impossible for humans to manage.
Nevertheless, even with the progress made in AI, people continue to play a crucial role in the workflow. Several middle stages still require hands-on software tasks. "A person is still needed to bring everything together," Niu mentioned.
According to Niu, experts continue to handle top-tier strategic planning, taking charge of key choices like determining priority targets and deciding whether to refine current compounds or create completely new ones from the ground up.
Artificial Intelligence for Science (AI4S) is transforming the life sciences through the incorporation of AI into scientific studies. Demis Hassabis and John Jumper from Google DeepMind received the 2024 Nobel Prize in Chemistry regarding AlphaFold, an artificial intelligence system designed to forecast protein structures.
In late December, Generate:Biomedicines, a company based in the United States, revealed intentions to initiate two worldwide Phase 3 studies examining an artificial intelligence-designed antibody for the treatment of asthma.
Chinese AI4S firms are also experiencing substantial growth. A startup backed by Baidu BioMap claimed it has surpassed Alphabet's subsidiary Google DeepMind's AlphaFold. in the monetization of artificial intelligence foundational models within pharmaceutical research.
Beijing-based start-up DP Technology has recently concluded its Series C fundraising round. , raising US$114 million.
AI-powered drug discovery firm Insilico Medicine which follows a business approach comparable to MindRank, was listed in Hong Kong last December. According to Niu, MindRank intended to apply for an initial public offering in Hong Kong this year, aiming for a listing in 2027.
Although there have been advancements, Niu cautioned that currently, AI might not transform the life sciences as significantly as it has done in other areas.
Niu stated, 'The main challenge is that the trial-and-error process takes way too long,' expressing concern over an issue affecting both major biopharma companies and startups, as shown by the extended validation periods within the sector.
Niu stated, 'For AI4S to create significant influence in the field of life sciences, we must continue with an extended process of testing and assessment.'
More Articles from SCMP
The real estate sector continues to pose challenges for Beijing's economic goals.
A Hong Kong company targeted by U.S. sanctions during the Venezuela operation was discovered to have used a false address.
15 people stranded as Hong Kong Ocean Park's 'Wild Twister' attraction malfunctions once more
"Cease undermining the Venezuelan government," says China, urging the U.S. to promptly free detained Maduro
The article was first published on the South China Morning Post (www.scmp.com), a top-tier news outlet covering stories about China and Asia.
© 2026. South China Morning Post Publishers Ltd. All rights reserved.
Posting Komentar