The "New" Chemistry: Open Source

Here's a plea for help posted by Derek Lowe on his blog In the Pipeline:

One of the people I met this past weekend was Matt Todd, chemistry professor at the University of Sydney. We talked about a project his lab is working on, and I wanted to help call attention to it.

They're working on praziquantel, also known as PZQ or Biltricide, which is used to cure schistosomiasis in the tropics. It's on the WHO's list of essential medicines for this reason. But PZQ is used now as a racemate, and this is one of those cases where everyone would be better off with a single enantiomer - not least, because the active enantiomer is significantly easier for patients to stand than the racemic mixture. Problem is, there's no cheap enantioselective synthesis or resolution.

So what Todd's group has done is crowdsourced the problem. Here's the page to start with, where they lay out the current synthetic difficulties - right now, those include enantioselective Pictet-Spengler catalysts and help with the resolution of a key intermediate. They were in need of chiral HPLC conditions, but that problem has recently been solved. I'd like to ask the chemists in the crowd here to take a look, because it wouldn't surprise me if one of us had some ideas that could help. Don't leave your suggestions here, though; do it over at their pages so it's all in one place.

This sort of thing is an excellent fit with open-source models for doing science: it's all pro bono, and the more eyes that take a look at the situation, the better the chance that a solution will emerge. I don't think it's getting the publicity it deserves. And no, in case anyone's wondering, I don't think that this is how we're all going to end up discovering drugs. Figuring out how to do this for large commercial projects tends to bring on frantic hand-waving. But in cases like this - specific problems where there's no chance for profit to push things along - I think it can work well. It makes a lot more sense than that stuff I was linking to last week!

Go to Derek Lowe's post to pick up the embedded links, especially to pick up the key link "here's the page to start with".

I'm guessing this will work and be a success. Lots of chemists won't mind taking a look and if they have an idea, put their two bits in. Open source works in the software world despite some rather high hurdles. When you want to help, it requires

a heck of a lot more than just "take a look". It is a substantial effort to access the code base, take on a task, develop, test, and submit the change/fix/enhancement.

I'm not a chemist, but looking at the requesting page, a lot of what Matt Todd asks for requires just a few minutes if you know the answer. This is so much easier than open source software, I don't see how it can fail to attract the help it needs.

In the past I've helped by choosing one of the collaborative distributed computing projects (see list here) and spent time helping there. I worked on "Folding@Home". I ended up spending many, many hours. But they set it up as a kind of "game" and contest where you could see your "score" versus other collaborators. It got the competitive juices going and I ended up spending more time that I intended. Ah... but it was for a "good cause".