Collaborative Chemistry

In the Wall Street Journal today (section B), nestled under an article on Star Wars aim to be a Box -office behemoth, was a small article by Denise Roland on drug giants. This described  the news that AstraZeneca and Sanofi announced  they will each share 210,000 of their compounds from their 2-3 million libraries. This is unprecedented because of the scale of the sharing. To me it sends a message of desperation. each company has run out of avenues to pursue so lets borrow our competitor compounds.  That would be like in formula 1 if Ferrari swapped cars with Mercedes but kept the same drivers. They can no longer innovate themselves. They may as well all become virtual pharmaceutical companies, leveraging compounds that are commercially available globally or from academic collaborators and do not do any work in house. Frankly that may be more profitable. In recent days we saw Boston Pharmaceuticals headed by ex Sanofi CEO  Christopher Viehbacher emerge as a virtual pharma shop.

While the AZ spokesman suggested the compound collection is a means to an end. The reality is that many of those compounds in the 2 million collection were probably purchased or made for them by chemists at other companies. Some of these compounds see the light of day in patents and publications, so  are the companies really sharing secrets? The AZ spokesman suggests their differentiating factor is the ‘biologists and scientists’ working for them. (I always believed that biologists are scientists – but seems that we are now a different species). Actually, as reported previously, the most important aspect of the screening of such compounds may be how they are dispensed and as we know AZ definitely use acoustic dispensing. I am not sure what Sanofi use, but small things like this could have a big impact on screening success. The article does not touch on the quality of such screenng data and also misses that many companies use virtual screening of massive libraries before focusing on compounds they can buy from vendors.

In the past we have proposed drug companies shared not just their compounds but their ADME/Tox data or models. I think that would have far more impact on productivity than sharing 10% of their cast off compounds. Like the GSK sharing of malaria and TB data in the past, this is just a small token effort, when what is really needed is something more transformational. For example all drug companies could collaborate to share all their libraries, or just their shelved compounds. They could all share their ADME/Tox data or models. The technologies exist for secure and selective sharing of data. Such examples of collaborations already exist and are funded by the NIH, EC, BMGF as well as others. I believe these will be the template which pharmas will ultimately use for all their research (eventually). In the meantime, I think that smaller, nimbler virtual pharma type companies may have the upper hand while large pharma try to stay relevant by slowly adapting and collaborating with each other.