06 May 2008

Open Source Drug Discovery

There's something utterly perverse about the way new drugs are developed. Pharmaceutical companies spend hundreds of millions - sometimes billions - of Euros investigating vast numbers of new compounds in the hope that they might treat a particular disease. If they find one that works, they then have to test it extensively for side-effects and the rest. Moreover, most of the negative knowledge they acquire - what doesn't work - is wilfully thrown away, since it represents "competitive" information.

But how about turning things on their head? Instead of trying millions of new substances for one disease, how about experimenting with the tens of thousands of known, safe medicines in the public domain on thousands of diseases? Like this:

The Johns Hopkins Clinical Compound Screening Initiative is an open-source effort to collect and index more than 10,000 known medications and determine which of them are also effective against hundreds of low-profile, Third World killers, such as Chagas disease, cholera and leprosy. The library will function something like a Wikipedia of drug discovery, where scientists around the world can contribute to the database and even provide samples or screen drugs themselves, thereby saving millions of dollars on R&D.

This could save millions of lives. Just one problem: nobody gets obscenely rich in the process....


Unknown said...

What you are suggesting is already being done. Companies are investigating "repurposing" known drugs to look for alternative therapeutic uses. Others are looking for uses of combinations of known drugs. In my opinion, though, this is likely to be a difficult process, given that drugs have evolved to be selective (even with the caveat that all possible activities could not possibly have been tested for).

Glyn Moody said...

Sure. What interests me here is the fact that it is being done in an open way; it will be interesting to see how much of an advantage that is in this context.