Artificial intelligence was used to identify an antibiotic that can kill a bacteria type responsible for many drug resistant infections.

Researchers at Massachusetts Institute of Technology (MIT) and McMaster University identified the new antibiotic from a library of almost 7,000 possible drug compounds.

Researchers used a machine learning model that they had previously trained to determine whether a chemical would inhibit the growth acinetobacter.

James Collins from MIT’s Institute for Medical Engineering and Science and Department of Biological Engineering said that the research supports the notion that “AI could significantly accelerate and expand the search for novel antibiotics”.

Advertisement

I’m thrilled that this research shows we can use AI in order to combat pathogens that are problematic, such as Acinetobacter Baumannii.

Acinetobacter Baumannii can cause pneumonia, meningitis, and other serious illnesses.

Jonathan Stokes is an assistant professor at McMaster University of biochemistry and medical sciences. He said that acinetobacter could survive for long periods on hospital equipment and doorknobs. It can also pick up resistance genes to antibiotics from its surroundings.

It’s common to find Acinetobacter baumannii that is resistant to almost every antibiotic.

Read More:

Elon Musk claims artificial intelligence isn’t necessary for anything

AI was the key to a breakthrough that enabled a paralysed man walk again

Weekly Podcast: Can AI take your job?

Researchers hope to develop compounds that will help patients with other drug-resistant infections using their model.

Nature Chemical Biology published their research.

Artificial intelligence can also be used to fight breast cancer. It helps scientists create a model that could predict if an aggressive branch will spread.

The AI model detects changes to the lymphnodes of women with breast cancer triple negative. Breast cancer is often found in the lymphnodes below the arm, on the same side. Patients in this case are likely needing more intensive treatment.

The development will give doctors a “new tool to help prevent secondary breast cancer”, according to Dr Anita Grigoriadis. She led the research in the Breast Cancer Now Unit of King’s College London.

She said: “By showing that changes in lymph nodes can predict whether triple-negative breast cancer will spread we have built on our increasing knowledge of the role that immunity can play in understanding patient prognosis.”

Researchers tested their AI model using more than 5,000 nodes donated to biobanks by 345 patients. The model then was able to determine the likelihood of spreading breast cancer by analysing immune response.

Triple negative breast cancer accounts for approximately 25% of breast cancer death in the UK.