The roots of bioinformatics

By Grant Jacobs 02/08/2010

I’d like to bring to wider attention an article series, The Roots of Bioinformatics.

The roots of prairie plants (Source: Wikimedia Commons.)
The roots of prairie plants (Source: Wikimedia Commons.)

Hold up.

Before experimental biologists run away, you’re some of the people I’d like this to reach!

At a recent bioinformatics meeting I attended, a senior experimental biologist from overseas commented to the effect that it was hard for bioinformatics researchers to build their own careers, to publish their own papers. My take on his comment was that he was referring to the tendency of experimental research to treat the bioinformatics as a service, seconding the work or linking to service centres, rather than treating bioinformatics researchers as collaborators, stand-alone groups in their own right.

I’m not about to argue the ins and outs of that here, but to draw attention to understanding the origins of bioinformatics, and hence it’s deeper alliance with biology.

One senior scientist at that meeting showered praise on Lincoln Stein for ’his pioneering work in bioinformatics’ (I’m paraphrasing here). I gritted my teeth and bit my tongue. Stein certainly contributed in a fine way to the human genome project, but he was no pioneer of bioinformatics.*

I thought it very telling that a senior biologist would think that.

It made me think that some (still) think of bioinformatics as a young field that emerged as a consequence of the service needs of the genome projects. I worry that an overly service-oriented view of bioinformatics prevails because of this.**

It’s a topic I touched on in The mythology of bioinformatics. My article doesn’t attempt to give a history of the field, as this series in PloS Computational Biology I am introducing here does, but briefly points to early work whose that could clearly be recognised today as being bioinformatics, and highlights other issues that I felt at that time were leading to misunderstanding.

By contrast, the new series in PloS Computational Biology is squarely aimed at giving an anecdotal narrative history of the field; another approach to better understanding bioinformatics today and what it has to offer to biology.

David Searl’s editorial introduces the Roots of Bioinformatics series, starting with a little light philosophy on the dual roles of tools and knowledge, bioinformatics combining the two.

The  second paper in the series, Russell Doolittle’s The Roots of Bioinformatics in Protein Evolution is now available. I like the approach he takes, working from the biological issues of the day, looking at how issues from the pre-electronic approach to protein biology lead to protein bioinformatics.

I would like to like to see this series more widely read, beyond preaching to the choir, as it is being published within a specialist journal.

My only worry is that the series will prove too patch-meal or specialist for those outside of the field. Hagen’s article might be useful to some (free PDF copy).

It will be interesting to see what the rest of the series brings, and I welcome that attention is being paid to the deeper origins of the field.


* All I am saying is that the field has older origins than his excellent contributions.

** I’m not saying that services don’t have a role, they certainly do, but that there is more to the field than just services.


Searls, D. (2010). The Roots of Bioinformatics PLoS Computational Biology, 6 (6) DOI: 10.1371/journal.pcbi.1000809 (open access)

Doolittle, R. (2010). The Roots of Bioinformatics in Protein Evolution PLoS Computational Biology, 6 (7) DOI: 10.1371/journal.pcbi.1000875 (open access)

Hagen JB (2000). The origins of bioinformatics. Nature reviews. Genetics, 1 (3), 231-6 PMID: 11252753 (subscription, but free PDF copy available on-line)

Other articles on Code for life:

Loops to tie a knot in proteins?

Who has the most bioinformatics scientists?

Testing common ancestry to all modern-day life

Consumer brain-computer interface

Forgetting older science

0 Responses to “The roots of bioinformatics”

  • Bioinformatics is really no different to astatistics when it comes to the tension between research and service (actually, much of bioinformatics is really no different to statistics at all). Statistics has developed into a mature field with its own journals that only specialists read, along with a career structure, internal squabbles, and everythign else a field needs to be considered mature. I don’t see why bioinformatics should be any different.

    I guess having articles looking back on teh past written by (presumably – I don’t know the authors) old farts is a step in that direction.

  • The man with the invisible comments strikes again 🙂

    Hi Bob,

    Bioinformatics is really no different to astatistics when it comes to the tension between research and service

    There will be parallels with other areas; I wasn’t meaning to imply that this was unique to bioinformatics!

    actually, much of bioinformatics is really no different to statistics at all

    Some of it, yes, but if you hold to this too strongly (or rather, too broadly) it becomes a misconception IMHO. This type of work is probably more visible as it probably gets better press – ? (e.g. large genetics/genomics studies are more likely to end up in Nature than, say, an exploration of the details of a protein family).

    As you know (I’m just writing this to keep the conversation up), a fair amount of bioinformatics is — ultimately — based on the properties of molecules, which is derived from chemistry and/or physics. Traditionally that’s particularly true for work on proteins, but it applies elsewhere too. IMHO there should be a lot more physics/chemical-based work than there is. (Expanding on this is a long story, better placed in a blog post I think.)

  • Hi Grant,
    You’re missing out on some well-deserved buzz! @razoralign and @biomednews both credited me via Twitter with your interesting blog post. As I work in advertising in NYC and not bioinformatics in NZ, I hope they get their sources fixed to correctly credit you for your work.
    Grant (@grantjacobs)

  • Thanks for letting me know.

    It makes me think that I need a credits line with my contact details spelt out! (I have also been meaning to get a new twitter account for some time as the one, but never seem to find the time to. The one I have uses my consultancy business name.)

  • Now, where are those comments? 🙂

    Anyhow, it’s true that many wet-bench biologists see bioinformatics as a means to another end. Heck, I do. I try not to.

    I don’t think I’ll ever forget the look of surprise – not necessarily pleased surprised, even – when we wanted to include a mathematician on a recent paper. It is so far removed from his field, and probably so useless to his career advancement, that he was indifferent. I suppose this is a good thing, in that he has adequate publishing opportunities elsewhere. Nonetheless, we insisted, so he takes responsibility for his contribution to the pursuit of truth. I think we’re all ok with the arrangement.

    Anyhow, nothing is better for collaboration than face-time. This is why I’m thrilled my research center is making active efforts to bring a critical mass of bioinformatics researchers into the fold. It’s a little funny, because you’d think that a collaboration would be easy to establish at a distance – my needs and interests are many! ie. where do heart- or eye-determining transcription factors bind in the human genome? what do the targets have in common? – because of it being computer-based. But the human interaction appears to be primordial.

  • Y’know when you see in hindsight that a word could be read more than one way? 🙂

    I meant the comments I cited within the article not the replies to the article! e.g. the bit starting a senior experimental biologist from overseas commented to the effect

    I agree face-to-face can help collaboration and given a choice I prefer it, if only for the usual human reasons 🙂 (It’s one of the reasons I leave myself open to working away from my base.)

    I think it depends on the nature of the work and the amount of effort you can put into contacting over a distance and other things too, though. I have had projects where only periodic face-to-face contact was needed (with on-going telephone/email/etc contact).

  • I do agree with the analogy of bioinformatics been somewhat like statistics. In some cases we are a service and in some cases we do basic research on improving our own tools and algorithms. The comparison is fuzzy, isn’t everything. What I do most days is not that different from business intelligence.

    I see a problem with many of us been shy and retiring or poor communicators or both. Then we complain that people treat us like service providers and only give grants for collaborations. My advice, get out of your office/ cave find someone to work with and lead the collaboration your self. Basically “Nut Up!”.

    Hardly anyone gets grants for single discipline research anymore. Lots of people get lots of money for inter-disciplinary work that needs informatics. There is no reason you can’t lead this. In fact if you do chances are the data will be a whole lot better.

    If you don’t like being the service, become the leader.