AI succesfully selects antibiotics candidate

MIT researches have trained an AI system on databases of antibiotics molecules. Then they have used the system to select from a set of candidates the most promising molecules that were structurally different from existing antibiotics. The system selected 8 molecules, from which the team finally picked one. The molecule was named halicin, after the 2001 - Space odissey computer, HAL. Halicin has proved effective in the lab against multiple baceria, some of which are highly resistant against current antibiotics. Although the system only helped to narrow down the selection to concentrate efforts on some candidates, it looks like a great help to increase the efficiency of the antibiotics research.