An artificial intelligence platform developed by Insilico Medicine has identified 12 novel compounds with potent senolytic properties — the ability to selectively eliminate senescent "zombie" cells that accumulate with age and drive chronic disease.
Using deep learning algorithms trained on molecular databases of over 500 million compounds, the AI system predicted molecular structures optimized for targeting senescent cells while leaving healthy cells unharmed. Laboratory testing confirmed that seven of the 12 candidates showed significant senolytic activity in human cell cultures.
"Traditional drug discovery would have taken decades to screen this many candidates," said Dr. Alex Zhavoronkov, CEO of Insilico Medicine. "Our AI identified these compounds in just 46 days, dramatically accelerating the path to potential therapies."
Senescent cells are a major driver of age-related diseases including osteoarthritis, atherosclerosis, pulmonary fibrosis, and neurodegeneration. While existing senolytics such as dasatinib plus quercetin have shown promise, they can cause significant side effects. The newly identified compounds appear to be more selective and better tolerated.
The three most promising candidates are now advancing to preclinical testing in animal models, with human trials anticipated to begin in late 2027.