MindRank Doses First Patient in Phase III MOBILE Trial of AI-Designed Oral GLP-1 Agonist MDR-001

MindRank, a clinical-stage AI-empowered drug discovery company, announced on 25 February 2026 that the first patient has been dosed in the Phase III clinical trial of MDR-001, its AI-designed oral small-molecule GLP-1 receptor agonist for obesity and type 2 diabetes in China. The study, MOBILE, is a multicenter, randomized, double-blind, placebo-controlled trial evaluating the long-term efficacy and safety of MDR-001 in approximately 750 patients across 50 clinical centers in China, led by Professor Linong Ji of Peking University People's Hospital. The program progressed from project initiation to U.S. IND clearance in 19 months and advanced to Phase III within approximately 4.5 years. In a completed Phase IIb trial, MDR-001 demonstrated clinically meaningful efficacy, favorable tolerability and improvements in cardiometabolic parameters. “Dosing the first patient in Phase III represents a major milestone for MDR-001 and for AI-enabled drug development,” said Zhangming Niu, Founder and Chief Executive Officer of MindRank. MDR-001 is a biased, selective oral GLP-1 receptor agonist discovered and optimized using MindRank's proprietary AI-enabled drug discovery platform, Molecule Pro. The team used the PharmKG biological knowledge graph to integrate literature and biological pathway analysis, and Molecule Dance for protein dynamic simulation modeling and biological validation iterations. This work analyzed the GPCR protein-small molecule binding conformation and signal transduction mechanisms and allowed quantification of specific pathway activities, including cAMP agonism and biased beta-arrestin 2 recruitment, to guide the design of a highly selective GLP-1 receptor agonist. Molecule Pro's generative AI and ADMET prediction modules, combined with wet-lab iterative feedback, enabled systematic prediction and optimization of the candidate's structural druggability and safety. MindRank describes Molecule Pro as an AI-driven platform that integrates nearly one hundred original algorithmic models, driven by deep learning, generative AI and large language models.