News · 2026-09-07

Insilico’s AI-designed IPF drug shifts six proteomic ageing clocks in a trial reanalysis

A reanalysis of an idiopathic pulmonary fibrosis trial found that all six proteomic ageing clocks moved in the younger direction in treated arms receiving rentosertib, an AI-designed drug candidate from Insilico Medicine. The finding is a clinical-trial ageing signal in a disease cohort, not proof that the drug rejuvenates healthy people or delivers a longevity benefit.

Key facts

• Nature Biotechnology published the study on September 7.
• It reanalysed a randomized, double-blind, placebo-controlled phase 2a IPF trial at 21 sites in China.
• Seventy-one patients were randomized; 42 contributed longitudinal proteomics for this analysis.
• The clearest cross-clock agreement appeared around week four in the 30 mg twice-daily arm.

Rentosertib is described in the paper as an AI-designed TNIK inhibitor from Insilico’s drug-development program. The study asks a question different from the earlier lung-function result: do patterns of proteins in blood look more like those of a younger person after treatment? To answer it, the authors applied six “proteomic clocks.” These are statistical models trained to estimate biological-age-related signals from many proteins, not literal instruments that measure a person’s age.

All six clocks pointed in the younger direction in treated arms. The 30 mg twice-daily arm showed the widest agreement and the clearest signal near week four. The result is meaningful because agreement across six methods is harder to dismiss than a movement in one proprietary score. It is not a table of six independent clinical outcomes, though. The clocks observe partly overlapping biological information, and their outputs are proxy measures.

The paper’s internal comparison provides its most useful restraint. The earlier lung-function signal was strongest in the 60 mg once-daily arm, rather than lining up simply with the 30 mg twice-daily arm’s cross-clock result. If a lower estimated biological age were merely a restatement of better lung function, those patterns would be expected to match more neatly. They do not, which is interesting, but it does not identify the mechanism.

A simple analogy helps. A proteomic clock is like a panel of weather instruments that infer a season from temperature, humidity and vegetation. If all show spring-like conditions, that is evidence about the environment. It does not prove the calendar has been turned back, nor explain which mechanism caused the shift. Here, the environment is an IPF patient undergoing treatment; disease state itself can change the protein signals that age clocks read.

Insilico’s press release uses stronger language, highlighting up to six years of reversal in one clock at week four. The paper is more cautious. It says the trial cannot fully disentangle disease improvement from ageing modulation, does not claim a longevity benefit, and calls for validation in healthy volunteers. That caveat is not fine print; it is the boundary of the result.

The concrete anchor is 42: only 42 participants supplied the longitudinal proteomics used for this new analysis. That is a real clinical dataset, but not a large, definitive ageing trial. The authors’ statement that the six clocks “support simultaneous geroprotective assessment” is better read as a proposal for how future trials might track multiple biological effects, not a declaration that ageing has been reversed.

Why it matters: AI-enabled drug discovery will increasingly generate claims at the boundary between molecule design, biomarkers and clinical benefit. The durable evidence chain is still the same—independent replication, patient-relevant endpoints, dose-response clarity and studies that separate disease recovery from a general ageing effect.

Primary source, verified: read the paper →

Ground Truth. (2026, September 7). Insilico’s AI-designed IPF drug shifts six proteomic ageing clocks in a trial reanalysis. Ground Truth. https://groundtruth.day/news/insilico-rentosertib-proteomic-aging-clocks-ipf.html

@misc{groundtruth:insilico-rentosertib-proteomic-aging-clocks-ipf,
  title  = {Insilico’s AI-designed IPF drug shifts six proteomic ageing clocks in a trial reanalysis},
  author = {{Ground Truth}},
  year   = {2026},
  month  = {sep},
  url    = {https://groundtruth.day/news/insilico-rentosertib-proteomic-aging-clocks-ipf.html}
}

Topics: biotech · drug-discovery · research · health · ai-for-science