AI Finds an Antibiotic to Combat Drug-Resistant Infections

Researchers from the Massachusetts Institute of Technology (MIT) and McMaster University have utilized artificial intelligence (AI) to discover a novel antibiotic that can combat drug-resistant infections. The AI, a branch of machine learning, was employed to identify a drug capable of combating drug-resistant infections caused by the Acinetobacter baumannii bacterium, which frequently appears in hospitals and can result in severe conditions such as pneumonia and meningitis. This bacterium is also the primary cause of infections associated with trauma in soldiers.

The researchers trained the AI to assess whether a chemical compound could impede the growth of A. baumannii. This led to the identification of a new drug from the pool of 7,000 compounds. This study was published in the journal Nature Chemical Biology and highlights how machine learning methods facilitate the swift exploration of chemical space, thereby enhancing the likelihood of discovering novel antibacterial molecules.

According to Jonathan Stokes, a former MIT postdoc, A. baumannii is highly adaptable and capable of surviving on hospital surfaces and acquiring antibiotic-resistance genes. He further emphasized the increasing prevalence of resistant strains of A. baumannii. This recent breakthrough in identifying a new antibiotic reinforces the valuable role of AI in expediting the drug discovery process.

Further, the research team trained the AI to recognize chemical structures that could hinder the growth of Escherichia coli bacteria. After screening over 100 million compounds, they identified a molecule called salicin, which exhibited efficacy against E. coli and against multiple other bacterial species, as well.

In animal trials conducted on mice, the team demonstrated that the newly discovered drug, named abaucin, effectively treated wound infections caused by A. baumannii while also exhibiting efficacy against several other drug-resistant infections.