A drug designed with the help of artificial intelligence made the blood profiles of people with a serious lung disease look biologically younger, according to a new study published in Nature Biotechnology. The finding emerged from an analysis of protein patterns in blood samples, where six independent biological aging clocks all detected a shift in the same direction - toward a lower predicted biological age - in patients who received the drug compared to those who did not.
The drug, called rentosertib, was originally developed to treat idiopathic pulmonary fibrosis (IPF), a condition that progressively scars the lungs and makes breathing difficult. IPF typically affects older adults and involves inflammation, cell damage, and problems with tissue repair - biological processes that overlap with changes seen during aging. Researchers from Insilico Medicine, Harvard Medical School, and other institutions decided to test whether a drug targeting lung disease might also influence markers linked to aging itself.
How the analysis worked
The team returned to blood samples from an earlier clinical trial of rentosertib involving 71 people with IPF. For the new analysis, they examined samples from 42 participants with an average age of 67. Researchers measured nearly 3,000 proteins in the blood - molecules that can carry recognizable traces of inflammation, tissue damage, energy use, and aging.
They then ran those protein patterns through six different biological aging clocks. Some clocks were built to estimate a person's chronological age. Others were designed to predict health and mortality risks. Using six models matters because no single clock is a perfect measure of biological aging. A result from one model could be influenced by the specific proteins it examines or the way its calculations work. When several independently developed clocks detect a similar change, the result is less likely to be a quirk of one model.
What the clocks detected
All six clocks detected younger-looking protein patterns in people treated with rentosertib. The placebo group showed little change, with some clocks even suggesting an increase in biological age. The clearest results emerged after four weeks. In the 60-milligram once-daily group, clocks designed to estimate chronological age showed reductions ranging from 2.71 to 3.46 years. Other clocks and dosing groups produced different figures, including much larger estimates from some exploratory models designed to examine individual organs. The researchers did not identify a single number as the overall effect.
The important finding was the direction of the change. "Despite being built in different ways, all six clocks detected younger-looking protein patterns in people who received the drug," the researchers reported. Rentosertib also altered the levels of 326 proteins, compared with only two in the placebo group. Some were linked to lung scarring and tissue repair. Others were involved in inflammation, energy use, cellular stress, and the behavior of old or damaged cells. Many of these protein changes continued through the 12-week study, but the apparent movement toward a younger biological age stopped increasing after week four and reached a plateau. The researchers do not yet know why. The body may have adapted to the drug, a different dosing schedule may be needed, or the aging clocks may have reached the limit of what they could detect during such a short study.
What the study cannot say
The results come with important limitations. Every participant had IPF. If rentosertib reduced fibrosis or otherwise improved their lung disease, that improvement alone could have made their blood look biologically younger. The clocks may have detected an improvement in the disease rather than a change in aging itself. The study could not fully separate those two possibilities. The analysis also included only 42 people and lasted 12 weeks. It did not demonstrate that participants had become younger, would remain healthier for longer, or would live longer.
Several authors work for Insilico Medicine, the company developing rentosertib. Lead author Alex Zhavoronkov is the company's founder and co-CEO. The findings will need to be tested in larger and longer studies, including trials involving people who do not have IPF. For researchers working at the intersection of drug development and aging biology, the study points to a future where drugs developed to treat diseases of old age might help scientists discover whether wider biological processes linked to aging can be changed as well.
Why this matters for science and research professionals
The study demonstrates a practical application of biological aging clocks as exploratory endpoints in clinical trials - a methodological signal that could influence how future drug studies are designed. For research scientists, the multi-clock approach offers a template for reducing model-dependent bias when assessing whether a compound affects aging-related biomarkers. The work also highlights the growing role of AI in drug discovery, an area where structured learning paths like the AI Learning Path for Research Scientists can help professionals build relevant skills. The broader AI for Science & Research field continues to expand as computational methods intersect with experimental biology, and studies like this one - published in Nature Biotechnology - will shape how research teams evaluate and validate AI-assisted discoveries.
Your membership also unlocks:
