Facebook has developed an artificial intelligence approach that it says can quickly predict how drugs will interact within cells and help identify combinations of drugs that could be used to treat cancer, New Scientist reports.
As researchers led by Facebook's David Lopez-Paz report in a preprint posted to BioRxiv, they developed an approach dubbed Compositional Perturbation Autoencode (CPA) to model and predict single-cell changes that occur with exposure to different drugs and doses, across cell types. According to New Scientist, the researchers found that their CPA approach could predict cell responses with about 90 percent accuracy and say this approach could hasten researchers' abilities to develop new treatments, including for cancer.
The US National Cancer Institute's Eytan Ruppin, who was not part of the study, tells it that the study is "important" first step, but that more testing is needed, especially as CPA predicts RNA changes that occur in the cell following treatment, but not whether that treatment leads the cell to die. "We have cured cancer one hundred times in salines and mouse models. They have shown nothing at all that is relevant to patients," he adds at New Scientist.