MindWalk Advances AI-Designed GLP-1 Program With Dual-Pathway Regimen for Longevity
MindWalk, formerly known as ImmunoPrecise Antibodies (IPA), has expanded its AI-driven GLP-1 therapeutics program by introducing a dual-pathway regimen aimed at aging and longevity. The approach combines the company’s AI-optimized GLP-1 receptor agonists with a proprietary companion therapeutic that targets a second, undisclosed pathway linked to cellular resilience and healthy aging. According to the company, this represents a first-in-class regimen with no currently marketed equivalents. Intellectual property filings for the therapeutic and regimen design have been initiated.
The regimen is intended for preventive, chronic use, with the aim of maintaining GLP-1 efficacy while broadening effects to include healthspan outcomes such as stress moderation and metabolic balance. Preclinical design includes exploration of dosing, scheduling, and biomarker endpoints relevant to aging biology. MindWalk positions this work as part of a broader strategy to apply its Bio-Native AI platform across therapeutic areas, including oncology, immunology, and vaccine design.
The program builds on MindWalk’s earlier work designing GLP-1 analogs in silico, which reportedly matched or exceeded semaglutide in receptor activation in preclinical assays. Using its LensAI platform, powered by the HYFT discovery engine, the company identified conserved biological patterns connecting GLP-1 signaling with metabolic and stress-response processes, while also mapping a complementary regulatory node relevant to longevity biology. The companion therapeutic was designed in silico and is exclusive to MindWalk’s pipeline.
MindWalk’s platform has also previously been applied to vaccine discovery. Earlier this year the company, when still operating as IPA, reported an entirely in silico identification of a conserved epitope across all four dengue virus serotypes using HYFT and LensAI.
HYFT is MindWalk’s patented system for encoding conserved sequence–structure–function relationships as universal biological “fingerprints.” By capturing these patterns, HYFT enables comparative reasoning across pathways and supports the identification of new biological mechanisms.
LensAI, built on HYFT, integrates sequence modeling, structural predictions, and knowledge-graph reasoning to translate these fingerprints into actionable insights. The platform has been applied to epitope mapping, de novo molecular design, and in silico vaccine development, and is linked to a wet-lab system that validates candidates under a closed design–build–test–learn loop. According to the company, this integration has supported more than 15 therapeutic candidates reaching clinical development.
Topic: AI in Bio
