WHY THIS MATTERS IN BRIEF
We can create vaccines faster than ever before, but human trials still take a long time, this could help accelerate drug approvals.
Interested in the Exponential Future? Join our XPotential Community, future proof yourself with courses from our XPotential Academy, connect, watch a keynote, or browse my blog.
Today, as we all struggle and strain with the increasing stresses associated with the current global coronavirus pandemic, COVID-19, Artificial Intelligence (AI) is being used to help accelerate the discovery of a vaccine for the deadly disease, as well as other diseases with some vaccines now reaching human trials. While there are now several plausible candidates in the wings one of the biggest hurdles companies trying to develop vaccines have to overcome is proving that they are safe – and in today’s world that means hundreds of thousands of trial vaccine doses, hundreds of thousands of volunteers, and months and months of waiting for the trial results.
Now though, just like they found a way to accelerate the development of new vaccines, cutting the time down from a decade or more to now just six months, researchers are trying to use AI to help them predict which drugs will be safe to use before companies even begin their trials.
When you take a medication, you want to know precisely what it does, which is why pharmaceutical companies go through extensive testing to ensure that you do, and this is where Metabolic Translator, a new AI computational tool that predicts metabolites, the products of interactions between small molecules like drugs and enzymes in the body, could help improve the process.
The new tool takes advantage of deep learning methods and the availability of massive reaction datasets to give developers a broad picture of what a drug will do. The method is unconstrained by rules that companies use to determine metabolic reactions, opening a path to new discoveries.
“When you’re trying to determine if a compound is a potential drug, you have to check for toxicity,” says Lydia Kavraki, a professor of computer science at Rice University, as well as co-author of the new paper in Chemical Science.
“You want to confirm that it does what it should, but you also want to know what else might happen,” she says.
The researchers trained Metabolite Translator to predict metabolites through any enzyme, but measured its success against the existing rules-based methods that are focused on the enzymes in the liver. These enzymes are responsible for detoxifying and eliminating xenobiotics, like drugs, pesticides, and pollutants. However, metabolites can form through other enzymes as well.
“Our bodies are networks of chemical reactions,” says graduate student and lead author Eleni Litsa. “They have enzymes that act upon chemicals and may break or form bonds that change their structures into something that could be toxic, or cause other complications. Existing methodologies focus on the liver because most xenobiotic compounds are metabolized there. With our work, we’re trying to capture human metabolism in general.
“The safety of a drug does not depend only on the drug itself but also on the metabolites that can be formed when the drug is processed in the body,” Litsa says.
The rise of machine learning architectures that operate on structured data, such as chemical molecules, make the work possible, she says.
Transformer was first introduced in 2017 as a sequence translation method that has found wide use in language translation and is based on SMILES, short for “Simplified Molecular-Input Line-Entry System,” a notation method that uses plain text rather than diagrams to represent chemical molecules.
“What we’re doing is exactly the same as translating a language, like English to German,” Litsa says.
Due to the lack of experimental data, the lab used transfer learning to develop Metabolite Translator. They first pre-trained a Transformer model on 900,000 known chemical reactions and then fine-tuned it with data on human metabolic transformations.
The researchers compared Metabolite Translator results with those from several other predictive techniques by analysing known SMILES sequences of 65 drugs and 179 metabolizing enzymes.
Though they trained Metabolite Translator on a general dataset not specific to drugs, it performed as well as commonly used rule-based methods that have been specifically developed for drugs. But it also identified enzymes not commonly involved in drug metabolism and not found by existing methods.
“We have a system that can predict equally well with rule-based systems, and we didn’t put any rules in our system that require manual work and expert knowledge,” Kavraki says. “Using a machine learning based method, we are training a system to understand human metabolism without the need for explicitly encoding this knowledge in the form of rules. This work would not have been possible two years ago.”
Source: Rice University