Not surprisingly the scramble to find treatments for COVID-19 is making productive use of AI. This week Fernanda Foertter, formerly of Nvidia and now a consultant for BioTeam, posted an article reprising BioTeam work meant to be a warm-up exercise for a machine learning webinar. Unexpectedly, the exercise identified a possible new COVID-19 drug candidate (montelukast) that is currently used to treat asthma. The model ranked montelukast higher than other more prominent anti-viral targets.
“As we did our work, our lead scientist had just taken their asthma medication, montelukast, so, out of pure curiosity, we used its molecular representation as the first input to play with the framework,” wrote Foertter. Whether the particular compound will turn out to be important against COVID-19 is a very separate question, but the brief post by Foertter digs into ML issues such as database quality, model bias, and the value of serendipity.
As explained by Foertter in her article (Leveraging AI to Hunt For Potential Treatments: A Covid-19 Example):
“In preparation for a webinar on April 29th we attempted to replicate the work done by Jonathan Stokes, et. al. at the Broad Institute in Cambridge, MA. They developed a model to predict new antibiotics from known drugs. Stokes leveraged data from the Drug Repurposing Hub to demonstrate how Halicin, a drug that showed poor results for the treatment of diabetes, could act as an antibacterial molecule.
“The ML framework Dr. Stokes’ team used ChemProp, was developed by The MIT Computer Science and Artificial Intelligence Laboratory and uses ensemble models to predict molecular properties using a method called message passing neural network. The framework can be trained on datasets containing molecules with known property values. These models can then be used to predict properties for new molecules.”
“The framework repository includes a few pre-built models, two of which were trained on drugs known to be active on a COV protein called 3-cytokine-like protease (3CLpro). The target, 3CLpro, is a non-structural protein that is a key enzyme in the viral life cycle of the COV involved in the maturation and activation of key viral replication proteins. Finding a drug that disrupts this protease would interfere with the viruses’ ability to replicate”
Pure chance led to montelukast being included in the study. Without going into details the team looked at many compounds. A summary table from the article is below.
Foertter emphasized, “While we are not offering scientific insight into this problem here, we do demonstrate how biomedical research organizations and pharmaceutical companies might be able to speed the discovery of treatment options using ML. These results are preliminary at best and need to be thoroughly explored and peer reviewed before any conclusions or medically-relevant actions can be taken. This exercise is about ML, not COV.”
Link to BioTeam article (Leveraging AI to Hunt For Potential Treatments: A Covid-19 Example): https://bioteam.net/2020/04/leveraging-ai-to-hunt-for-potential-treatments-a-covid-19-example/