San Antonio scientists use machine learning to identify potential treatments for deadly viruses

A team of San Antonio-based biomedical researchers trained a machine learning algorithm to identify more than two dozen viable treatments for diseases caused by zoonotic pathogens that can jump from animal hosts to infect humans. Scientists from Southwest Research Institute (SwRI), The University of Texas at San Antonio (UTSA) and Texas Biomedical Research Institute (Texas Biomed) used SwRI-developed Rhodium™ software to study bat-borne Nipah and Hendra henipaviruses, which are endemic to some parts of the world and cause particularly lethal infections in humans.

Through the collaboration, researchers mapped the protein structure of the measles virus, which is in the same family of viruses as henipaviruses. With measles as a blueprint, Rhodium virtually screened and ranked compounds for corresponding structures and binding effectiveness. Out of 40 million compounds, Rhodium identified 30 potentially viable viral inhibitors for Nipah and Hendra. Although the research focused on antiviral treatments for henipaviruses, any broad-spectrum therapeutic that's developed could potentially treat related viruses, including measles.

"The results suggest that machine learning can rapidly identify antiviral candidates for highly pathogenic viruses that are difficult to study due to space limitations and biosafety constraints," said Dr. Jonathan Bohmann, a staff scientist at SwRI, who presented these findings at Hendra@30 Henipavirus International Conference in Melbourne, Australia. "Our algorithms allow us to make the best use of resources to deliver a 'short list' of potential treatments for further testing."

This Department of Defense research is funded by the Peer-Reviewed Medical Research Program (PRMRP), under the Congressionally Directed Medical Research Programs (CDMRP), and opens the door to finding treatments for Nipah and Hendra. According to the World Health organization, 40-75% of people infected with these diseases die.

Henipaviruses are lethal pathogens. They're endemic to animal populations in Asia and Australia, but spillover events to livestock and humans occur regularly on a seasonal basis, which is concerning due to the pandemic potential."

Dr. Jonathan Bohmann, staff scientist at SwRI

Studying such infectious diseases requires adherence to strict safety standards and access to a BSL-4 rated high-containment laboratory. By virtually screening compounds, researchers save time and resources.

"Our work highlights the power of collaborative, multidisciplinary research by our San Antonio institutions to bring together a comprehensive and cohesive strategy for developing novel anti-viral drug candidates," Dr. Stanton McHardy, professor at The University of Texas at San Antonio and director of the Medicinal Chemistry and Synthesis Core Facility at the Center for Innovative Drug Discovery.

"Rhodium does a very good job of sifting out compounds that are toxic and finding effective disease inhibitors," said Dr. Olena Shtanko, assistant professor at Texas Biomed, who collaborated with McHardy and Bohmann by evaluating the effectiveness of the antiviral compounds identified by Rhodium. "We've made quite a lot of progress in a short amount of time, but more research is needed."

SwRI's Pharmaceutical and Bioengineering Department provides FDA-inspected facilities that meet Current Good Manufacturing practice standards.