petermr's blog

A Scientist and the Web


When does open science work?

It’s funny how things turn out in the blogosphere. I’d posted about how ludicrous copyright on dead scientists’ work (Copyright madness – story 2) was and expected some comment from the librarian community. Silence (there’s still time to comment!). I got a brief exclamation of horror from BlackKnight and to check that this wasn’t spam visited his blog. and I saw the Green Fluorescent Wow! My comments about this example of immediate Open Notebook Science has turned into a thread on when and whether to publish results on blogs. Here’s Black Knight:

Whee. I checked Technorati this evening, as you do (seeing as the bastard spammers have destroyed the usefulness of trackbacks), and discovered that yesterday’s post was spreading ripples in the blogospheric pond. It came first to the attention of Peter Murray-Rust, who has a thing (a good thing! — I hasten to add) about open access and open science in general, and thence to the open science community itself, in the shape of Cameron Neylon.

Funnily enough, Neil Saunders then picked it up from the OpenWetWare people, and I do some digging and find that not only did Neil do a DPhil but that he is now in Bostjan Kobe‘s lab. Bostjan is a long-time collaborator of my previous boss and from meeting him at Lorne he seems to have quite forgiven me for not going to work for him when I had the chance (or forgotten about it).

Brisbane, bleh.

So, anyway, it turns out that I had previously made contact with OpenWetWare, and talked about them over a year ago. Which just goes to show that (a) it’s a small world and (b) incest is more fun than you’d think.

But all that is not really what I wanted to write about now. The OpenWetWare (have you any idea how difficult it is to type that?) project is a laudable effort to promote collaboration within the life sciences. And this is cool, but then I realize that the devil is in the details.

Share my methods? Yeah! Put in some technical detail? Yea–hang on.

For sure, the ‘Green Fluorescent Wow!’ experiment (HT to Peter) was pretty simple and straightforward: An easy cloning experiment with a slight cleverness in choice of reagents, no IP and nothing particularly smart. But I’ve got other experiments underway that are clever, and potentially very exciting.

So can I write on my weblog about them? And how much detail can I give? If I say “My protein seems to do something odd to cell-motility”, is that an elegant sufficiency of detail? Surely people will get bored with generalizations, but am I right to worry, as one of our PIs does, that I might compromise my project by posting too much detail? Should I really be posting pictures of cells that are doing odd things?

It’s not a case of “Can I trust you bastards not to steal my work?” but balancing the ideal of ‘open source science’ with the need to publish before anyone else. I have responsibilities — to the boss and to my cow-orkers —, but I also want to share the fun and joy and heartache of this vocation.

So it’s all a little bit confusing, really. I want to bounce when experiments work, and scream and shout when I have a ‘little technical difficulty’; but how much can I say without compromising stuff? Seriously, I have lots I want to write about, but am not sure whether I should.

Comments (Cameron Neylon)

I think this is the real key to the whole thing. Will it compromise your work/career/future happiness? If everyone shared and was honest then it should work. What we need is some game theory/evolutionary biology person to tell us how many people it takes before we can support freeloaders.

But I agree it takes a shift in the way science works for people, especially postdocs, to be ready to risk making their knowledge base available. It would be absolutely key for people to get credit, and citations in some form, for making protocols/data available.

  • I don’t think anyone is suggesting that all science everywhere be automatically Open (e.g. there is – as yet – no Richard Stallman of Open Notebook Science. At at obvious level we have industrially sponsored projects in our group and we are required (and quite happy to be required) to check all new discoveries with the sponsors.
  • It’s very domain-dependent. In some areas such as maths and physics the idea can be the whole thing. A few seconds – such as the Watson-Crick DNA model or the Franklin data can communicate the whole message in a few seconds. But most science is made up of tedious work, much of which doesn’t.
  • It’s often thought that the “idea” is the most important thing. And sometimes it is. But most ideas don’t work out . Who has not had the grey-haired community (I’m one) telling them that it won’t work? Are they are sometimes right. So in many cases the credit goes to someone who has stuck doggedly at making a half-baked idea work.
  • There are, however, many cases where “your idea” has already been tried elsewhere and failed, and the reasons for its failure are documented and possibly understood. An awful lot of this happens at bars in conferences. “We’re trying to see how protein X might interact with Y” (not giving too much away). “Have you made Y?”. “No, we’re trying A’s method.” “Oh, we tried that for 2 years and we couldn’t get it to work; and A also published Z which didn’t work either. So we distrust stuff from that lab”.  If this is not disinformation then it’s very valuable and could save wasted years of work. Remember that the unproductive work has to be put on the balance sheet as well.
  • So – as Cameron says – it’s a game. You estimate the value of releasing your idea to the value of not releasing it. Either could be positive or negative.
But there are other pluses to publishing work on the Web:
  • You make a reputation. I’m not hiring green fluorescent rats but if I were I would recognise the applicant.
  • Your contacts may be genuine collaborators, not competitors. So, for example, if I were interested in a multidisciplinary programme on malaria including medicinal chemistry I would certainly wish to make contact with Jean-Claude Bradley. Indeed if I were interested in an open programme of screening compounds (as is the NIH) I would also make contact.

So “I’ve got something novel – I’m looking for collaborators” will become more common in the electronic era. Some of this will be public – others (as we heard at scifoo) will be mediated though brokers in private. Perhaps we’ll see embargo periods – publish the science into a closed arena for a few months with a requirement that it then becomes public. Ideas and novel science are too valuable to be allowed to decay.

2 Responses to “When does open science work?”

  1. Making new contacts should be one of the biggest motivations.
    For anyone looking for collaborators, a new job or a postdoc, it is ridiculously obvious how useful the blogosphere can be (just ask Bora now at PLoS ONE). I understand the paranoia about publication in traditional journals and I think the answer is to find a middle ground. Start by delineating one small project or sub-project to make public, preferably Open Notebook. People got used to the idea of discussing “unpublished” results at conferences – this is no different in philosophy.

  2. [...] Recent Comments Cameron Neylon on The Southampton E-lab blog notebook – Part 1 AimsJean-Claude Bradley on The Southampton E-lab blog notebook – Part 1 AimsUnilever Centre for Molecular Informatics, Cambridge – petermr’s blog » Blog Archive » When does open science work? on A great example of ‘fun’Cameron Neylon on A great example of ‘fun’Black Knight on A great example of ‘fun’ Subscribe Search for: [...]

Leave a Reply