Last week, the Food and Drug Administration approved the first treatment for Covid-19. The drug is called Veklury, although most people refer to it by its scientific name, remdesivir. But what very few people knew is that the recipe for manufacturing the drug has been hiding more or less in plain sight—at absolutely no cost—on the online code hosting platform GitHub.
Of course, one would need the hardware to recreate the almost 400-atom molecule formula. But one would imagine that Gilead, the pharmaceutical company that manufactures Veklury, would like to keep that recipe secret, considering it has brought in $873 million in revenue this year so far. With the announcement that Veklury would be ready for international distribution by the end of October, one might expect that revenue to be only a fraction of what it could reap.
(The FDA’s move to approve remdesivir puzzled some in the scientific community, who find its results at treating COVID-19-related effects to be mediocre at best.)
Standing in Gilead’s way are people like Lee Cronin, a chemist at the University of Glasgow. He and his team have developed software that translates a chemist’s words into recipes for molecules that robots understand. This led to his design of a robotic chemist he calls a “chemputer” that produces chemicals from XDL programs, like the one for remdesivir.
A CNBC story on Cronin inadvertently provided the information about remdesivir’s availability on GitHub. The story describes Cronin’s dream of researchers being able to produce and share molecules in a standard system “as easily as they email and print PDFs”—like an “ebook reader of chemistry.”
His lab first gave a look at the machine they designed, which is capable of producing multiple molecules, last year, and now they’ve progressed toward digitizing chemistry in an easier way to program with the chemputer. The software they made translates academic papers into programs for chemputers that researchers can edit without having to code.
Key to Cronin’s chemputer is the chemical description language or XDL, which gives instructions for the robotic chemist machine to know what to do. His team developed its own software that scans a chemical recipe—like remdesivir—and puts the instructions into XDL, which then directs the chemputer to execute actions with standard heaters and test tubes.
Cronin is working toward making this system compatible with other chemistry robots, so any researcher with their own machines can use it. In this quest, Cronin’s team is not alone. Dozens of groups around the world in academia and industry are working to this safer and more accessible means of digitizing chemistry. This sense of promoting scientific collaboration could lead to many benefits to be used in everything from space exploration to sharing formulas with developing countries to give them better access to medications without having to pay a fortune to pharmaceutical giants.
Just don’t tell Gilead.