News
AI-Designed Drug Lowered Patients' Biological Age in Insilico Medicine Trial
An analysis published in Nature Biotechnology shows that rentosertib, an experimental drug designed by Insilico Medicine's AI platform, lowered biological aging markers in patients treated for a rare lung disease.
Insilico Medicine has published an analysis in Nature Biotechnology showing that rentosertib, a drug designed from scratch by its artificial intelligence platform, lowered patients' biological age as measured by six independent aging clocks based on blood proteins. It is the first evidence this extensive that a drug candidate discovered entirely by generative AI can have a measurable effect on aging processes, not just on the disease it was designed to treat.
How the Drug Was Designed
Rentosertib, also known as ISM001-055, is a TNIK protein inhibitor developed by Insilico Medicine's Pharma.AI platform. The company chose TNIK as its therapeutic target because the protein is involved in six different mechanisms considered hallmarks of cellular aging. The drug was originally tested as a treatment for idiopathic pulmonary fibrosis, a rare and progressive disease in which lung tissue gradually scars.
The phase IIa results of that trial were published in Nature Medicine in June 2025 and showed improved respiratory parameters in treated patients. For the new analysis, researchers returned to the same 71 phase 2a participants and analyzed blood samples from 42 of them who had additionally agreed to detailed proteomic testing at further time points during therapy.
Six Clocks, One Result
The team measured 2,841 proteins in blood serum using the Olink panel, and each profile was scored with six independently built proteomic aging clocks, including ProtAge, two variants of OrganAge, PAC, ipfP3GPT, and PAOPAC. These clocks compare a given patient's protein profile against a reference database of 55,319 profiles from the UK Biobank, allowing biological age to be estimated independently of chronological metrics.
Across all six clocks, patients treated with rentosertib tended to show a reversal of biological age relative to the placebo group, even though the tools were built in different labs using different methodologies. The largest effect was observed at week 4 of therapy in the group taking 30 mg twice daily, where the clocks indicated patients were biologically younger by 2.7 to 3.5 years, with one clock showing a difference of up to 6 years. In the group taking the higher 60 mg once-daily dose, lung vital capacity increased by an average of 98.4 milliliters, while it declined by 20.3 milliliters in the placebo group.
What convinces me is not the size of the effect, but the agreement, the models share neither features nor training data - Michael Levitt, 2013 Nobel laureate in Chemistry
Scientists' Caveats
The authors and commentators caution that proteomic aging clocks are not a universally accepted clinical standard, and relying solely on biological age as a measure of treatment efficacy could lead to flawed medical decisions. The analysis described here is also a secondary study of existing blood samples from the phase 2a trial, not a new clinical experiment designed from the start to measure aging.
The consistency across the clocks demonstrates that rentosertib's effect on aging-related proteomic signals is a robust biological phenomenon - Prof. Jing-Dong Jackie Han
What's Next for the Drug
Rentosertib has already been cleared to enter phase III clinical trials in China for idiopathic pulmonary fibrosis. Insilico describes it as the first drug candidate discovered and designed entirely by generative artificial intelligence to advance this far in the regulatory process, making it a test of whether AI-designed drugs actually move through successive clinical trial phases faster and more cheaply than drugs developed with traditional methods.
Insilico Medicine, founded by Alex Zhavoronkov, went public on the Hong Kong stock exchange on December 30, 2025, under the ticker 03696.HK. In the first half of 2026, the company reported roughly $106 million in revenue, a 287 percent year-over-year increase, and the total value of contracts it signed this year reached around $7.3 billion. Over the first nine months of 2026, the company nominated nine new drug candidates for further development.
The result fits into a broader trend of investment in AI aimed at longevity. Examples cited alongside this analysis include OpenAI chief Sam Altman, who invested $180 million in a startup focused on cellular rejuvenation, alongside other projects involving blood plasma transfusion and its substitutes. The growing number of such initiatives makes it easy to overrate individual lab results, which is why scientists stress the importance of independent replication.
For Poland's biotech market and investors tracking the AI-in-pharma sector, the result shows that artificial intelligence models, from molecule design to the analysis of proteomic data from clinical trials, are starting to meaningfully shorten certain stages of drug development. Still, a years-long process separates confirmation of an effect in a proteomic analysis from regulatory drug approval, and rentosertib remains, for now, a phase III candidate rather than an approved drug.
