Hundreds of potential COVID-19 drug candidates identified thanks to machine learning

RIVERSIDE, Calif. — The scientists of the world have yet to agree on a sole, ideal drug to treat the novel coronavirus. According to a new study, however, there certainly isn’t a shortage of possibilities. Researchers at the University of California, Riverside have identified hundreds of COVID-19 drug candidates through the use of machine learning.

“There is an urgent need to identify effective drugs that treat or prevent COVID-19,” says research leader Anandasankar Ray, a professor of molecular, cell, and systems biology, in a release. “We have developed a drug discovery pipeline that identified several candidates.”

This “drug discovery pipeline,” as professor Ray calls it, is in actuality a computational AI-connected algorithm capable of teaching itself to predict and detect patterns and activity over time through trial and error.

With a vaccine still months away, identifying a drug that proves truly effective against coronavirus could save countless lives.

“As a result, drug candidate pipelines, such as the one we developed, are extremely important to pursue as a first step toward systematic discovery of new drugs for treating COVID-19,” professor Ray explains. “Existing FDA-approved drugs that target one or more human proteins important for viral entry and replication are currently high priority for repurposing as new COVID-19 drugs. The demand is high for additional drugs or small molecules that can interfere with both entry and replication of SARS-CoV-2 in the body. Our drug discovery pipeline can help.”

How artificial intelligence can give rise to COVID-19 treatments

What makes this complex algorithm tick? Joel Kowalewski, a graduate student in professor Ray’s lab, used a small grouping of ligands (molecules) associated with 65 human proteins known to come in contact with SARS-CoV-2 proteins. For each of the 65 proteins, a new machine learning model was created.

“These models are trained to identify new small molecule inhibitors and activators — the ligands — simply from their 3-D structures,” Kowalewski notes.

All that allowed researchers to develop a database of chemicals featuring structures predicted to interact with the 65 proteins.

“The 65 protein targets are quite diverse and are implicated in many additional diseases as well, including cancers,” Kowalewski says. “Apart from drug-repurposing efforts ongoing against these targets, we were also interested in identifying novel chemicals that are currently not well studied.”

In all, more than 10 million (from a database of over 200 million) commercially available small molecules were screened by the machine learning models. Then, among the molecules that “hit” for any one of the 65 proteins, researchers looked for compounds that have already been approved by the FDA. Any potentially toxic compounds were weeded out by the machine learning models.

This process is what allowed them to identify drug candidates with the highest potential for fighting the coronavirus. Some chemicals are even predicted to neutralize at least two of the 65 protein targets.

“Compounds I am most excited to pursue are those predicted to be volatile, setting up the unusual possibility of inhaled therapeutics,” professor Ray says.

Machine learning shows invaluable potential in modern medicine

“Historically, disease treatments become increasingly more complex as we develop a better understanding of the disease and how individual genetic variability contributes to the progression and severity of symptoms,” Kowalewski notes. “Machine learning approaches like ours can play a role in anticipating the evolving treatment landscape by providing researchers with additional possibilities for further study. While the approach crucially depends on experimental data, virtual screening may help researchers ask new questions or find new insight.”

This newly-developed computational strategy represents a big improvement over older ways of analyzing large assortments of chemicals simultaneously.

“Our database can serve as a resource for rapidly identifying and testing novel, safe treatment strategies for COVID-19 and other diseases where the same 65 target proteins are relevant,” professor Ray concludes. “While the COVID-19 pandemic was what motivated us, we expect our predictions from more than 10 million chemicals will accelerate drug discovery in the fight against not only COVID-19 but also a number of other diseases.”

The study is published in Heliyon.

Like studies? Follow us on Facebook!

Leave a Reply

Your email address will not be published. Required fields are marked *