Employing artificial intelligence (AI) in the realm of “drug discovery” has opened up a new, more efficient, and cost-effective approach to identify potential medications. A group of researchers has used the technology to identify three promising candidates for senolytic drugs, which have the potential to slow the aging process and mitigate age-related diseases.
According to an articlewritten by Vanessa Smer-Barreto, a scientist from the University of Edinburgh, a group of researchers fed an AI model with examples of known senolytics and non-senolytics, and the model successfully distinguishing between the two, predicting new potential senolytics.
Senolytic drugs function by eliminating senescent cells, also dubbed “zombie” cells. These are metabolically active cells that are unable to replicate due to DNA damage. While in some cases, their inability to replicate prevents the spread of damage, an accumulation of these cells can lead to a host of diseases, including Type 2 diabetes, COVID, pulmonary fibrosis, osteoarthritis, and cancer.
Does this mean AI can cure cancer? Not yet—but it’s one step in the right direction.
Lab-based studies on mice have demonstrated the potential of senolytics to alleviate these conditions by exterminating zombie cells and preserving healthy ones. But the operational costs are too high to be tested on humans at a large scale.
“It would be great to find more senolytics that can be used in a variety of diseases, but it takes ten to 20 years and billions of dollars for a drug to make it to the market,” Vanessa Smer-Barreto wrote.
But here is where AI does its magic. The researchers used the best-performing model on a set of 4,340 molecules and in just five minutes (remember this number), it delivered a list of 21 high-probability senolytic candidates of which three compounds (periplocin, oleandrin, and ginkgetin) demonstrated the ability to eliminate senescent cells while largely preserving the normal ones.
The advantages of the use of AI are evident. “If we had tested the original 4,340 molecules in the lab, it would have taken at least a few weeks of intensive work and £50,000 just to buy the compounds, not counting the cost of the experimental machinery and setup,” Smer-Barreto noted.
Just yesterday, Decrypt reported that a group of researchers published a paper about a protein language model named Ankh. Developed to understand inter-protein communication and discover new proteins, this research underscores the potential of AI in scientific innovation.
Stability AI, the company behind Stable Diffusion, revealed that it’s also working on an AI-powered technology to study and understand proteins.
The implications of integrating AI into traditional scientific disciplines are vast. By accelerating the pace of drug discovery, AI promises to propel the search for cures and treatments. The three newly-identified senolytics are currently undergoing testing in human lung tissue, with the next set of results anticipated in two years.