The approach is based on two popular AI techniques: generative racing networks and training enhancement .
News: A team of AI pharma startup Insilico Medicine, working with researchers at the University of Toronto, took 21 days to create 30,000 designs for molecules that target fibrosis-related protein (tissue scars). ) They synthesized six of these molecules in the lab, then tested two in cells, with the most promising one being tested in mice. The researchers concluded that it was potent against the protein and showed "drug-like qualities". In all, the process took only 46 days. The study was published in Nature Biotechnology this week.
Context: Getting a new drug on the market is very expensive and time-consuming: it can take 10 years and cost as much as $ 2.6 billion, with the vast majority failing at the testing stage, according to Center for Drug Research . No wonder there is so much work to do on AI to accelerate the process. DeepMind is among the companies researching pharmaceutical research as a potential future avenue for their algorithms.
A word of caution: Although the study seems promising, it is still a very proof of concept, We are far from creating drugs designed with intellectual intelligence, let alone sold to patients. We examined the problem in this article of our TR10 issue earlier this year.
Register here for our daily newsletter Download to get your dose of the latest industry-leading breaking news.