A machine learning approach enables generalizing and identifying subtle patterns in chemical structures to discover new classes of potential drugs, unlike the brute force method used by humans. The approach predicts entire chemical substructures that define hundreds of different chemicals, leading to the discovery of new potential antibiotics. After obtaining a list of compounds with high predicted antibiotic activity, the next step involves testing their effectiveness and toxicity. The study found two compounds sharing the same predicted substructure, defining a new class of antibiotics, and confirmed their effectiveness.
Casey is taking his newsletter Platformer off Substack, as criticism over the company’s handling of pro-Nazi content grows. Then, The Wall Street Journal spoke with witnesses who said that Elon Musk had used LSD, cocaine, ecstasy and psychedelic mushrooms, worrying some directors and board members of his companies. And finally, how researchers found a new class of antibiotics with the help of an artificial intelligence algorithm used to win the board game Go.
Today’s guests:
- Kirsten Grind, enterprise reporter for The Wall Street Journal
- Felix Wong, postdoctoral fellow at M.I.T. and co-founder of Integrated Biosciences
Additional Reading:
We want to hear from you. Email us at hardfork@nytimes.com. Find “Hard Fork” on YouTube and TikTok.