416-479-0074

Machine Learning and Human Health: Decoding New Antibiotics

Machine Learning and Human Health: Decoding New Antibiotics

antibiotics

Long-time readers of this newsletter can corroborate: We’re always interested in the development of AI through machine learning. We’ve seen computer bits of intelligence fool university students, teach English to Japanese schoolkids, name kittens, and sort Lego.
 
While it is fun to think about a computer dubbing a baby cat Snox Boops, how well does machine learning work with less frivolous data? Well, a team from MIT has found out, by challenging an AI to pore through thousands of pharmaceutical compounds and come up with a working antibiotic. And it has succeeded — unbelievably well.
 
“To find new antibiotics, the researchers first trained a ‘deep learning’ algorithm to identify the sorts of molecules that kill bacteria. To do this, they fed the program information on the atomic and molecular features of nearly 2,500 drugs and natural compounds, and how well or not the substance blocked the growth of the bug E. coli.
 
Once the algorithm had learned what molecular features made for good antibiotics, the scientists set it working on a library of more than 6,000 compounds under investigation for treating various human diseases. Rather than looking for any potential antimicrobials, the algorithm focused on compounds that looked effective but unlike existing antibiotics. This boosted the chances that the drugs would work in radical new ways that bugs had yet to develop resistance to.”
 
The AI found a stellar combo, which the researchers cheekily named “halicin,” after the meddling computer HAL 9000 in 2001: A Space Odyssey. But the antibiotic itself is far more helpful to humans than anything HAL was responsible for: In tests, it has cleared the bacterium behind tuberculosis and C. difficile, as well as a host of other, equally drug-resistant bugs.
 
Its creators are hoping to work with a non-profit or pharmaceutical company to bring halicin to the market in the near future. Until then, their concept proven, they will continue to throw molecules at their trusty AI — who knows what medical wonders will come out the other side!