AI Discovers Method of Killing Drug-Resistant Bacteria

One of the most prominent concerns that exist in the drug market today is the incidence of drug-resistant strains of bacteria. With more people starting antibiotic treatments and not completing them, the volume of drug-resistant cases is continually rising from year to year. Scientists at MIT have unveiled a new drug called Halicin that doesn’t work the same way that others on the market do. By changing the way that the compound interacts with bacteria, making it more difficult for drug-resistant strains of pathogens to avoid destruction. The new compound is named Halicin, in honor of HAL9000 from 2001, A Space Odyssey.

What Makes This Drug Different?

Most modern pharmaceutical compounds that enter testing take existing drugs and modify them slightly to make something that the bacteria isn’t aware of, The slight changes make it impossible for the bacteria to fend off the new drug until it gets used to it. This approach is usually why pharmaceutical companies refer to the development of new compounds as an arms race since they have to stay one step ahead of the bacteria. Halicin approaches the problem differently by attacking the electrochemical gradient that bacteria need to produce molecules that store their energy. The AI that developed Halicin was trained on a combination of existing artificial pharmaceutical compounds as well as naturally occurring ones that destroy bacteria. Armed with this knowledge, the AI sorted through a library of six thousand new compounds, and determined Halicin would be the most effective one for wiping out bacteria.

When Does it Come Out?

The discovery stage of a pharmaceutical compound is just the first step in the development of a drug. So far, researchers have used the drug in lab trials that show promise. It may be up to a decade before we can use this drug as consumers. The MIT team successfully repelled an infection of A. baumanii in mice. The news is heartening, as the pathogen is common in US soldiers posted of Iraq or Afghanistan. The AI has been put to work looking for new compounds and has presented a potential pool of 23 new compounds after scanning over 100 million molecules.

CloudWedge