The Massachusetts Institute of Technology (MIT) has developed an AI algo which helped to quickly define a new bug busting molecule.
Drug resistant bacteria are on the increase making some diseases impossible to treat with the usual antibiotics. The hunt was on for a solution.
Using a previously developed algo which had been trained to analyse and seek traits in particular molecules, those with a high probability of killing bacteria, the model was reassigned to look for molecules capable of killing E. coli. It was trained on 2,500 molecules which included almost 1,700 FDA approved drugs and a group of 800 natural products known for their diverse structures and broad spectrum bio-activities. Naturally, the perfect molecule would not have to cause toxicity to humans, so a different algo was used to screen out any harmful ones.
The new molecule was named after the computer in 2001 A Space Odyssey, (the one with the disembodied voice called Hal). Halicin has been tested against some really nasty bacteria such as Clostridium difficile, Acinetobacter baumannii and the old favourite TB, Halicin performed very well.
Just to exercise their muscles after finding Halicin, researchers applied the algo to another 100 million molecules from the ZINC15 database (which contains 1.5 billion chemical compounds) completing the task in just three days. They were happy to discover a further 23 ‘possibles’ which were also predicted to be safe for humans.
This is the second new drug found using AI reported on this site and I am sure you will agree with me…..the flood gates are beginning to open!