In what has been called the first time in the history of humanity, a powerful learning algorithm was used to study and analyze more than one hundred million chemical compounds in a matter of days. The algorithm found a compound that was able to kill 35 types of potentially deadly bacteria said, researchers.
The discovery is being hailed as one of the biggest push in the fight against drug-resistant bacteria. The AI’s potential is finally being harnessed to create drugs that can solve some fraction of the problem.
The researchers used what is called a Neural Network to analyze the structure of 2500 drugs and other compounds to find the ones that were the most likely to be an effective choice against an increasingly drug-tolerant E. Coli strain.
The new paper that carries the information about the methods used also carries pre-clinical results, which are a necessity when it comes to proving that an AI-discovered antibiotic is just as safe and likely to work than any other drug out there. The added speed of computers could help bring down the cost of new drugs in the future.
The use of AI in the medical area is still in its early stages, although the concept is not new. One of the more famous displays of AI superiority has been in the area of diagnosing breast cancer from mammograms. The learning algorithm deployed to predict breast cancer from a given mammogram was built by a team of international researchers. The algorithm was tested against six highly competent radiologists, and it was able to outperform every one of them.
The use of AI could help us fight the rising trend of antibiotic-resistant bacteria as learning algorithms work with no bias whatsoever. When our traditional knowledge and methods fail to get a solution to this problem, AI could step in and provide us with a fresh approach and possibly a solution.