This article is more than 1 year old
Machine-learning software scours database of already available drugs that could treat COVID-19 infections
Next up: Clinical trials to test the code's suggestions
Machine-learning algorithms may help pharmaceutical companies identify drugs that can be rehashed to treat the COVID-19 coronavirus in elderly patients, according to a study published in Nature Communications.
More than two million people worldwide have been killed by the human malware, and more than 100 million have caught the bio-nasty, according to official statistics. As governments race to vaccinate billions of people to stop the spread of the SARS-CoV-2 coronavirus, a group of researchers from MIT, Harvard University, and ETH Zurich are looking to old-fashioned medicines to treat infected patients.
Want to let an AI-powered doctor loose on patients? Try slapping a food-label-like sticker on it, says Uncle SamREAD MORE
“Making new drugs takes forever,” said Caroline Uhler, lead author of the paper, and an assistant professor focused on computer science and biology at MIT. “Really, the only expedient option is to repurpose existing drugs.”
Here’s where AI systems come in handy. Searching through large databases of available drugs and inspecting their pharmacological effects to identify those that could treat a COVID-19 infection is no easy feat. Massive amounts of data have to be carefully analysed to home in on the right drug candidates that will effectively treat COVID-19.
The researchers decided to focus on the fact that older people tend to suffer most from the virus because their lung tissue becomes stiffer with age. They built software that walked back from proteins and genes linked to both aging and coronavirus infections to their genetic root cause, and then searched for available drugs that tackle the effects of those genes. The idea being that those drugs may also lessen the blow of the virus, and help a patient recover.
Here's how MIT described the team's work:
First, they generated a large list of possible drugs using a machine-learning technique called an autoencoder. Next, they mapped the network of genes and proteins involved in both aging and SARS-CoV-2 infection. Finally, they used statistical algorithms to understand causality in that network, allowing them to pinpoint “upstream” genes that caused cascading effects throughout the network. In principle, drugs targeting those upstream genes and proteins should be promising candidates for clinical trials.
To generate an initial list of potential drugs, the team’s autoencoder relied on two key datasets of gene expression patterns. One dataset showed how expression in various cell types responded to a range of drugs already on the market, and the other showed how expression responded to infection with SARS-CoV-2. The autoencoder scoured the datasets to highlight drugs whose impacts on gene expression appeared to counteract the effects of SARS-CoV-2.
“By intersecting the resulting combined SARS-CoV-2 and aging interactome with the targets of the top-ranked FDA-approved drugs from the previous analysis, we identify serine/threonine and tyrosine kinases as potential drug targets for therapeutic interventions,” the group's paper stated.
All of this led to the scientists homing in on RIPK1, a protein that has been linked to cell death. It's hoped that drugs can alleviate the effects of COVID-19 by altering the function of RIPK1. Choosing compounds that affect the enzymes involved in the function of the protein could prevent symptoms like inflammation.
“In independent research, RIPK1 has been shown to bind to SARS-CoV-2 proteins and has also been found to be in an age-dependent module,” Uhler told The Register.
"In addition, RIPK1 belongs to an interesting family of proteins that can take on very different roles depending on which of its domains is activated: in particular, it can be a mediator of inflammation and cell survival, but it can also be a mediator of cell death potentially triggering tissue fibrosis
“We hypothesize that upon SARS-CoV-2 infection in older individuals the death pathways may be favored. Consistent with this, recent post-mortem lung tissue biopsies of SARS-CoV-2 human patients revealed a fibrotic epithelium and increased blood clotting. Going forward, it would be critical to experimentally test this hypothesis, for example by developing an organoid model that mimics the aged lung."
“Collectively, our results suggest 11 FDA-approved protein kinase inhibitors as candidate drugs for the repurposing against COVID-19 in the elderly, among which three drugs have been tested or are currently being tested in clinical trials around the world,” she concluded.
Uhler said the team plans to share their results with pharmaceutical companies, and won’t know how effective their predictive model is until elderly patients are enrolled in real clinical trials.
“Rigorous in vitro experiments as well as clinical trials are needed to validate the identified candidate drugs,” the paper added. ®