Some interesting news out of MIT - a study where a number of existing medicines were used to identify potential antibiotics. Seemed like a bit of a long shot approach as they identified features of “good antibiotics” and then looked for them. The candidate selected was effective in treating a number of multi-resistant infections and was originally designed for Diabetes…
http://news.mit.edu/2020/artificial-intelligence-identifies-new-antibiotic-0220
Jon - I posted a longer form review of the paper on Saturday, and highlights in a thread earlier with @mary.crowe that was already in the theme of drug-development and AI, if interested.
opengraphobject:[360523051229184 : http://news.mit.edu/2020/artificial-intelligence-identifies-new-antibiotic-0220 : title=“Artificial intelligence yields new antibiotic” : description=“Using a new machine-learning algorithm, MIT researchers have identified a powerful antibiotic that can kill a wide range of species of pathogenic bacteria, including some that are resistant to all known antibiotics.”]
Yes - link here - I have picked up the media end of it. Good to see that they have actually tested the compound in the lab to see that it works - as you say the algorithm isn’t so much the star as just another tool like automated assays…
opengraphobject:[360523051229184 : http://news.mit.edu/2020/artificial-intelligence-identifies-new-antibiotic-0220 : title=“Artificial intelligence yields new antibiotic” : description=“Using a new machine-learning algorithm, MIT researchers have identified a powerful antibiotic that can kill a wide range of species of pathogenic bacteria, including some that are resistant to all known antibiotics.”]