Swamidass: Artificial Intelligence Shakes Up Drug Discovery

My scientific work was recently highlighted in The Scientist.

The lead on aricle highlights work from my group:

Then, in 2018, graduate student Na Le Dang at Washington University in St. Louis hit upon a way to use artificial intelligence (AI)—specifically, a machine learning algorithm—to work out the possible biochemical pathways terbinafine takes when it is metabolized by the liver. Trained on large numbers of known metabolic pathways, the algorithm had learned the most likely outcomes when different types of molecules were broken down in the organ. With that information, it was able to identify what no human could: that the metabolism of terbinafine to TBF-A was a two-step process.

It’s really our ambition to automate as much as possible, so the chemists are just focusing on the much higher level, the difficult problems, the strategy.

—Adrian Schreyer, Exscientia

Two-step metabolites are much more difficult than direct metabolites to detect experimentally, which is likely why this potentially lethal outcome wasn’t flagged until after the product was on the market, says S. Joshua Swamidass, a physician scientist and computational biologist at Washington University and Dang’s supervisor. The discovery not only shed light on a long-standing biochemical mystery, but showed how the use of AI could more broadly aid drug discovery and development.

Given enough data, machine learning algorithms can identify patterns, and then use those patterns to make predictions or classify new data much faster than any human. “A lot of the questions that are really facing drug development teams are no longer the sorts of questions that people think that they can handle from just sorting through data in their heads,” Swamidass says. “There’s got to be some sort of systematic way of looking at large amounts of data . . . to answer questions and to get insight into how to do things.”

There is unquestionably a lot of hype around the potential of AI in drug discovery—Swamidass likens this period to the internet boom of the late 1990s. But many, such as Schreyer, are excited about the possibilities that this new technology offers, particularly when it comes to finding novel therapeutics for difficult-to-treat diseases.


You are actually DOING science. Congratulations and keep up the great work. Years from now, peoples lives are going to made better in part by your work. We (you) do science because it works. It really does makes a difference in people’s lives.


Couldn’t agree more. Human brains just aren’t set up to analyse massive data sets, and computers are an obvious tool for getting past that shortcoming.

I ran into a similar problem a while back when I was looking at gene and microRNA expression arrays. I tried to make sense of it, but beyond some obvious associations it turns into a massive spider web of interactions quite quickly. The real danger is constructing a hypothesis after you get the data. Sometimes it works, but it opens yourself up to false associations.


AI and machine learning are kind of general tags for a whole host of different applications and approaches. Currently, hypothesis driven work seems to pan out better than identifying correlations and hoping they’ll be interesting (e.g. ‘fishing expeditions’). But there’s going to be a lot of really nice applications either way. A bottleneck we’ll probably see is being able to evaluate generated hypothesis with experiments in the “wet” labs (e.g. “live biology”).

1 Like