New Superbug-Killing Antibiotic Discovered Using AI

Improve the News Foundation -

Facts

  • Artificial intelligence (AI) has helped a team of scientists discover a new antibiotic that can be lethal to a strain of deadly superbug. The US-Canadian team said that the finding may greatly speed up the process of discovering new treatments.1
  • On Thursday, the results from the team from McMaster University and the Massachusetts Institute of Technology to combat Acinetobacter baumannii — the superbug determined to be a "critical" threat by the World Health Organization — were published in Nature Chemical Biology.2
  • The result of this research was the robust yet experimental antibiotic called "abaucin." Before the new drug can be released to the market, it will require further testing.3
  • The bacteria can result in pneumonia, meningitis, and the infection of wounds. Acinetobacter baumannii can also acquire genes from other bacteria that are resistant to antibiotics.4
  • While broad-spectrum antibiotics kill many different kinds of bacteria at once, what the group of scientists is trying to develop is an antibiotic designed to work only against Acinetobacter baumannii to slow its rate of resistance.5

Sources: 1BBC News, 2Guardian, 3RNZ, 4Independent, and 5CNN.

Narratives

  • Narrative A, as provided by Brighter World. In search of new medical treatments, the old way was to test one chemical at a time. This is slow, labor-intensive, and costly. Modern computational approaches using artificial intelligence can assess the antibacterial properties of hundreds of millions, possibly billions, of molecules, enabling scientists to develop new drugs much faster and cheaper. AI has resulted in a true medical breakthrough here.
  • Narrative B, as provided by College of Computing & Informatics. Artificial intelligence may be used in healthcare systems to reduce costs and relieve clinicians of some workload, but its use also raises important concerns. While rendering some research jobs redundant is one of these, there are also other more major downsides too — the risk of missing data, excluding important social variables, and vulnerability to cyber attacks. AI must be used with caution and regulation.

Predictions