Insilico Medicine discovers IPF drug candidate ISM001-055 in 18 months for $2M using generative AI

Idiopathic pulmonary fibrosis (IPF) is a progressive fatal lung disease with limited treatment options. Traditional drug discovery requires years and hundreds of millions of dollars to identify a novel biological target and design a matching therapeutic molecule ready for human trials.

Insilico Medicine's AI platform identified a novel TNIK biological target linked to IPF, designed the small-molecule inhibitor ISM001-055, and completed preclinical studies entirely through generative AI-driven workflows. Candidate molecules were filtered through enzymatic assays, in vitro pharmacokinetic studies, and in vivo toxicology models. The full process was later documented step-by-step in a Nature Biotechnology paper.

The entire process from target identification to Phase 1 readiness took only 18 months at a cost of approximately $2 million. ISM001-055 subsequently demonstrated improvements in lung function in a 12-week placebo-controlled Phase 2a trial along with a favorable safety profile. The program has become a published benchmark for AI-driven drug discovery transparency.

• AI-driven discovery can compress target-identification-to-IND timelines from multiple years to under 18 months
• A ~$2M cost-to-Phase-1 represents a potential order-of-magnitude reduction versus traditional methods
• Publishing the full methodology in Nature Biotechnology helped establish scientific credibility for generative AI drug design