The approach is based on two popular AI techniques: generative adversarial networks and reinforcement learning .
News: A team from AI pharma startup Insilico Medicine, working with researchers at the University of Toronto, took 21 days to create 30,000 designs for molecules targeting a protein linked to fibrosis (tissue scarring) .) They synthesized six of these molecules in the laboratory and then tested two in cells, with the most promising one tested on mice. The researchers concluded that it was potent against the protein and showed "drug-like" qualities. All in all, the process only took 46 days. The research was published in Nature Biotechnology this week.
Context: Getting a new drug to market is hugely expensive and time-consuming: it can take 10 years and cost as much as $ 2.6 billion, while most fail at the testing stage, according to Tufts Center for the Study of Drug Development . No wonder then that there is so much work going on using AI to speed up the process. DeepMind is among the companies exploring pharmaceutical research as a potential future opportunity for its algorithms.
A word of caution: Although the research looks promising, it is still very much a proof of concept. . We are far from making AI-designed drugs, let alone sold to patients. We explored the case in this article from our TR10 edition earlier this year.
Sign up here for our daily newsletter. Download to get your dose of the latest must-read news from the world of new technology.