Mid UK Open Science Trip Report

I’m about half way through the UK trip that Cameron Neylon organized for me. So far it has proven to be a productive though exhausting Open Science fest. First I stopped by UKOLN at the University of Bath on August 29. Then the Nature Science Blogging conference in London followed by an Open Science Workshop at Southampton University then back to London for a talk at the Nature offices.

Cameron was with me at all these events and we even co-presented at Nature. It was a special pleasure to be able to report on the results we obtained in his lab just the previous day. Now a Google search for “boc-glycine solubility THF” pulls up the UsefulChem lab notebook page EXP207 as the first hit. What better example of Open Notebook Science in action?

The FriendFeed effect was in full force at a few of these events. People are finding it natural to microblog sessions by posting comments. Richard Grant, Cameron and I made good use of it at the Southampton conference. When a speaker started to present someone would start a thread and others would comment on that thread. It is interesting to see how FriendFeed is evolving within scientific communities. Some perceive that it is mainly to be used for ephemeral conversations. But why not bookmark conversations to document conferences for the longer term? A big problem is FF doesn’t generally make its way into the Google index but bookmarking should force that to happen for selected content.

The Southampton conference was also a great opportunity to discuss in detail the challenges and opportunities facing Open Science. If the ideas discussed there get fleshed out in document we may end up referring to a “Southampton Resolution on Open Science”. By the end I think we came to an agreement that a viable compromise might consist of sharing all raw data files associated with a paper after publication. This is a far cry from the ideal of Open Notebook Science, aiming for “no insider information”, that Cameron and I advocate. But it is a step in the right direction and it is not unreasonable to try to get a number of scientists on board.

I really enjoyed my conversation with David De Roure from MyExperiment at Southampton U. There is great potential for bringing cheminformatics workflows up to speed with bioinformatics. I agree with David that using ChemSpider web services on Taverna might bring us a long way in that direction. In organic chemistry automatically calculating masses and volumes of reagents based on moles and chemical identifiers like SMILES or InChIKeys certainly would come in handy for minimizing errors and planning time.

After the London Science Blogging conference, Cameron, Egon and I headed out to Peter Murray-Rust’s house for lunch and coding. We created a CMLreact file for one of the Ugi experiments that was part of our paper about to appear in JoVE (see here for Precedings version). This can be approached from many different levels of abstraction. We chose to focus on the equimolar 0.4M experiment run in triplicate and ignore the specific sequence or manner of reagent addition. As long as the abstractions point back to the original lab notebook pages with full details, simplifying the representations of reactions is not a problem. But there is certainly not a standard way of this as of yet using CML.

I'll put more up shortly, including my presentation in Manchester. Right now Duncan Hull is waiting for me to join him at the pub.....