Accessibility links

Breaking News

Robot Scientist Helps Design New Drugs

Robot Scientist Helps Design New Drugs
please wait

No media source currently available

0:00 0:01:12 0:00

This video shows Eve, the robot scientist. The voice in the sound track is Ross King, who heads the robotics team that developed Eve at the University of Manchester. In this video he describes how Robot Scientist Eve works to discover new drugs.

Robots have become partners in the laboratory. At the University of Manchester in England, a robotic scientist named Eve has discovered a compound known to have anti-cancer properties that might also be used to develop drug strategies against malaria and other tropical diseases.

Eve follows Adam, a robot scientist the University's Automation Laboratory built in 2009 to do research. Adam and Eve are not human look-alikes. They are boxy and about the size of a car. They have the machine quality of robots you might see in an automotive plant.

Team leader Ross King says Adam made a discovery about yeast, a fungus used in biology as a model for human cells.

“Adam hypothesized certain functions of genes within yeast and experimentally tested these hypothesizes and confirmed them," he said. "So it both hypothesized and confirmed new scientific knowledge.”

That was a first for a robot. Adam's success set the stage for Eve. Eve's science mission is focused on tropical and orphan diseases, which kill millions and infect millions more each year. King says these conditions are largely neglected because, on average, drug development is slow and costly. It can take a decade or more to get new medicines to market, and costs around $1 billion. Manufacturers are unlikely to get that investment back.

Eve learns from its success

King says the University of Manchester developed a cheap assay - a test that indicates whether or not a chemical is likely to be a good drug candidate - and then put Eve to work on it.

“How it works conventionally is you use robotics as well and you have a large collection of possible drugs," he said. "Against your assay you test every single compound. So you start at the beginning of your [compound] library and continue until the end, then stop. So it’s not a very intelligent process. The robotics doesn’t learn anything as it goes along, even if it’s tested a million compounds, it still doesn’t have any expectation of what will happen next when it tests a new compound."

That’s where Eve differs. The robot learns as it tests, King explained, “so that it can then [eliminate] the compounds which are unlikely to be good and only test the compounds which it has a high probability of being good.”

That’s how Eve unexpectedly discovered the compound that might fight malaria and other tropical diseases.

Making drug development faster, cheaper

King hopes to completely automate the process, with robots making assays and synthesizing new chemicals. He points out that while robots have become partners in the lab, humans are still very much in control.

“They do the strategic thinking about what’s important to tackle, and organizing the background knowledge and organizing the structure of the process," he said.

You can read more about Eve and Adam in this week's issue of the Royal Society journal, Interface.