AI Helps Scientists Develop New Antibiotic to Treat Deadly Superbug

Using artificial intelligence (AI), scientists from McMaster University and the Massachusetts Institute of Technology have found a new antibiotic that has the power to kill a deadly superbug.

This superbug, known as Acinetobacter baumannii, is considered a serious threat by the World Health Organization. In fact, it falls into the “critical” category among the organization’s list of “priority pathogens.” These pathogens are a collection of bacteria families that pose the highest risk to human health.

The scientists’ findings were published on Thursday in the journal Nature, Chemical Biology.

Resistance to drugs

The WHO has warned that these bacteria have a natural ability to develop resistance to treatment and can even transfer this resistance to other bacteria, making them resistant to drugs as well.

Acinetobacter baumannii is particularly concerning for hospitals, nursing homes, and patients who rely on ventilators and blood catheters, as well as those recovering from surgeries with open wounds.

AI discovered a new superbug-killing antibiotic 🤯

Scientists have used artificial intelligence (AI) to discover a new antibiotic that can kill a deadly species of superbug.

The AI helped narrow down thousands of potential chemicals to a handful that could be tested in the… pic.twitter.com/poz7bOIZIG

— AI Daily (@AlexAIDaily) May 26, 2023

This bacterium can survive on surfaces and shared equipment for extended periods of time, making it easily transmissible through contaminated hands. It is not limited to causing blood infections; it can also lead to urinary tract and lung infections.

The CDC also highlights that A. baumannii has the ability to “colonize” or inhabit a patient’s body without causing any infections or noticeable symptoms.

Training AI to discover new types of antibiotics

In this study, researchers utilized an advanced AI algorithm to screen numerous antibacterial molecules, aiming to uncover potential new types of antibiotics. Through this AI screening process, a fresh antibacterial compound was successfully identified and named abaucin.

Gary Liu, a graduate student from MacMaster University involved in the research, explained the process they undertook. They had access to a large dataset containing information about various chemicals and their effectiveness in killing bacteria.

Liu’s task was to train the model to determine whether new molecules possessed antibacterial properties or not. Essentially, the model’s role was to analyze and identify which molecules could potentially serve as antibiotics.

Liu highlighted the potential of AI in their research, explaining that it allows them to explore large areas of chemical space quickly.

Testing of new antibiotic

Once the AI model was trained, scientists utilized it to analyze 6,680 compounds that had not been previously encountered.

The analysis was completed within an hour and a half, yielding several hundred compounds. Out of these, 240 compounds underwent laboratory testing.

Moreover, through this testing, nine potential antibiotics were identified, including abaucin.

To further assess its efficacy, the new molecule was tested against A. baumannii in a wound infection model using mice.

Consequently, the results demonstrated that the molecule effectively suppressed the infection caused by the bacteria.