A graph neural network classifier screened over 7,000 plant compounds for molecules that interfere with the chemical signalling bacteria use to coordinate infection. Binding assays confirmed that 17 polyphenols, including pterostilbene and pyrogallol, attach to two of these signalling proteins in Salmonella Typhimurium. The compounds also changed the activity of those signalling genes and the bacteria's attachment to human gut cells grown in culture.
Why it is interesting: Machine learning used to find compounds rather than to predict outcomes, with the evidence stopping at laboratory work.
Cite this item
BibTeX
Free sign-up. A digest of the new papers twice a week, grouped by domain.