• krellor@kbin.social
    link
    fedilink
    arrow-up
    56
    ·
    11 months ago

    It sounds like they trained a classification model using 39,000 molecules with known reactivity to MRSA. The molecules are vectorized text representations of the structures. Once trained, they can run arbitrary molecules through the model and see which ones are predicted to have antibiotic properties, or at least MRSA reactivity.

    They likely fed in molecules from families of structures that seem likely to contain an antibiotic but are too numerous to manually test them all. They get a prediction of which ones are likely to have the properties they want, and then start the slow process of creating and testing the molecules in the lab.

    • xkforce@lemmy.world
      link
      fedilink
      English
      arrow-up
      22
      ·
      11 months ago

      I get what they did (its been something a lot of groups have been wanting to do for years) but I am curious what molecule specifically they found that worked especially well. i.e What does this thing look like? What is the new antibiotic’s mechanism of action? None of those latter details are discussed. Its something we can only guess at.

      • krellor@kbin.social
        link
        fedilink
        arrow-up
        18
        ·
        11 months ago

        It sounds like they are moving forward with clinical testing in partnership with a bio company, so I’m sure they withheld the information anticipating a patent. The results of this paper was the validation of the explainable AI model which identified candidate classes of compounds.