AI may have discovered the ultimate anti-aging compound
What's the story
Drug discovery is a demanding task, especially considering the time, research, and costs involved. In such a case, AI could provide a big boost and it has proven to do so. A team from the University of Edinburgh and the Spanish National Research Council have used AI to find three "most powerful" molecules that could slow down aging.
Detail
Newly found molecules help get rid of non-replicating senescent cells
The research group has found promising targets for senolytic drugs, which can slow down aging and prevent age-related diseases. These drugs work by destroying senescent cells—cells that are metabolically active but which have lost their ability to replicate. They are often called zombie cells. Senescent cells could turn out to be problematic. For instance, they can secrete inflammatory molecules that can affect neighboring cells.
Implication
Increased senescent cells are linked to diseases like cancer, diabetes
An increase in the number of senescent cells has been seen to have an effect on several diseases, including Type 2 diabetes, osteoarthritis, and even cancer. Studies carried out on mice have revealed that destroying senescent cells, by using senolytics, can improve these conditions. Senolytic drugs can eliminate zombie cells while keeping healthy cells alive at the same time.
Study
Machine learning models were trained to identify new drugs
In the study, the team trained machine learning models to identify new senolytic drug candidates. These models were fed on known examples of senolytics and non-senolytics. They could differentiate between the two and could also predict whether molecules that they were not trained on could be senolytic drugs. Researchers narrowed down on best-performing system based on the one which made the least errors.
Testing
The team tested the chosen AI model on 4,340 molecules
The team then tested the chosen AI model on 4,340 potential molecules to find which ones could be senolytic drugs. Out of this, the system picked 21 candidates and it gave results within five minutes. To put things into perspective, had this drug screening been done manually, it would have taken at least a few weeks, not to mention the experimental costs.
Results
The shortlisted drugs were tested on two types of cells
The 21 shortlisted molecules were further tested on healthy and senescent cells to find the best anti-aging drug. Of these potential drug candidates, three molecules, namely periplocin, ginkgetin, and oleandrin, were able to get rid of senescent cells, without affecting healthy cells. Upon further detailed analysis and experimentation, the team finally found that oleandrin was the most effective of the lot.
Future
The team is now testing them on lung tissues
The team intends to carry on with the investigation. Now that three senolytic drugs have been validated, the team has now moved to testing them on human lung tissue. However, it could take quite some time before we get to know how that experimentation went. The team hopes to release the results in about two years.