Dr. Rebecca Davis and Hunter Sturm
UM lab at the forefront of antibiotic drug development using artificial Intelligence (AI)
Researchers explore using AI to make better antibiotics
A team of UM researchers is using machine learning algorithms and artificial intelligence (AI) to improve and create new antibiotics at a faster rate to help address the growing problem of antibiotic resistance.
Chemistry professor Rebecca Davis and PhD candidate Hunter Sturm, from UM’s Davis Research Group, are applying machine learning algorithms and AI to help understand what is happening at each step of the antibiotic drug development process.
The team has developed a new understanding of machine learning algorithms that can help see the complexity of AI decision making through Explainable AI (XAI).
“We want to teach the computer how to show us what steps it is taking to arrive at a conclusion, rather than just assuming the steps it took are the correct ones,” says Sturm. “We know that AI is only as good as the information that is fed into it, so we want to remove the mystery and train the AI model on what it should look for and understand the steps it has taken.”
Davis and Sturm recently presented their work at the fall meeting of the American Chemical Society. They are among a small number of researchers in North America who are trying to build a greater understanding amongst the science community around this subset of AI and reduce skepticism around new methodologies.
“We collaborate with researchers at UM and around the world to make progress on drug development aided by computational and physical chemistry,” says Davis. “We definitely punch above our weight with this innovative work to help provide insight into how AI make decisions in order to help move forward with better antibiotics.”
As part of a larger international collaborative effort that combines the expertise of chemists (Davis, Manitoba), microbiologist (Cardona, Manitoba), bioinformaticians (Hu Lab, Western) and computer scientists (Friederich, KIT, Germany), they are working to develop methods to improve the antibiotic discovery pipeline.
This innovative research in drug development may lead to improved accuracy and less time to analyze data for large pharmaceutical companies. The team hopes this will dramatically reduce the time and cost of new drug development and improve the availability of antibiotics on the market.
Research at the University of Manitoba is partially supported by funding from the Government of Canada Research Support Fund.