A way has been found to make tumors age

To solve this problem, researchers have developed the SAMP-Score tool. It uses machine learning and the analysis of thousands of microscopic images of cells. The model was trained to recognize the characteristic signs of aging by the shape, structure and other visual features of cells — these are the so-called morphological profiles associated with aging. This approach makes it possible to distinguish a real stop of fission from false signals associated with toxicity or common variations.

Using this tool, scientists have tested more than 10,000 experimental chemical compounds. One of them, called QM5928, showed a particularly strong effect: it consistently caused aging in various types of cancer cells, including those that are resistant to standard drugs. At the same time, the cells did not die, but rather entered a state in which they stopped dividing.

It is especially important that QM5928 worked in tumors with high activity of the p16 protein, such as basal—like breast cancer, which often does not respond well to existing medications. The researchers also found that under the action of the compound, p16 moves into the cell nucleus — perhaps this is what helps to trigger the division block.

The creators emphasize that without SAMP-Score, it would be impossible to notice such subtle changes. The new tool shows how a combination of machine learning and high-precision visual analysis can accelerate the search for drugs that use natural aging mechanisms to fight tumors.

SAMP-Score is already available on GitHub and can become a universal tool for laboratories working on the development of new anti-cancer therapies, especially for cancers that are difficult to treat.