Friday, December 16, 2011

Open Source cancer research (must see video)

Four years ago I wrote up a passionate post about the importance of ODOSOS, and about a year ago written about how research can be open sourced, and how the Open Source Chemistry Development Kit (CDK) fits in. I am proud that the CDK is enabling so many researchers to do novel research! Google Scholar reports over 200 documents and Web of Science gets to over 150 for just the first CDK paper (see also GS vs WoS). That's serious impact. We're far from being fully comparable with existing commercial tools, which have a 40 year head start. But with important functionality still missing (e.g. E/Z stereochemistry), and about 150 open bug reports which needs looking into, we surely can need funding and help!

Anyway, this Monday we had the last of this years Stockholm Open Science meetings, and Carl Bärstad joined, whom is organizing TEDx talks here in Stockholm, and he pointed me to this must see video on open source cancer research:

I can highly recommend watching it, as it is both very insightful about cancer (it's the third mechanism I know now how cells remember state, after DNA methylation, microRNAs, and now plain, boring protein; I wonder when metabolite-sized molecules show up as cell-division surviving state preservatives; they will).

But, it also puts Open Science ideas to practice in drug discovery. Yeah, the even give the structure (and SMILES) of JQ1 (see PubChem):

Now, the cancer the started of with, is pancreatic cancer, which is what my mother died of 3 years ago, which provides a third reason for me to love this video!

Update: I have uploaded JQ1 and the charged variant on ChemSpider to Ambit2 as dataset 976496, which means that you can look up toxicity predictions with ToxPredict for JQ1 on this page. This is what ToxPredict looks like, just before I hit Run all:


  1. Awesome video, thanks for posting!

  2. Have watched this more than once and have just emailed Assistant Professor Bradner directly with a link included to this post.

    Great TED Talk - thanks for sharing, Egon.

  3. It's still frustrating that despite the advances of the genomics revolution this is still the most neglected form for cancer. I watched this video a few weeks ago and it made me wonder what's really holding back progress with respect to this disease. I jotted down some ideas about some of the recent trends in cancer biology. From an informatics perspective though, I think we're still lacking open source, web-based tools that let researchers screen compounds and generate curves. Beyond that, simply picking the right targets seems to be the biggest challenge. It's appalling the number of trials that fail because we didn't understand the biology well enough.