The great flagellum debate

Flagellae, the whip-like structures that many microorganisms use for propulsion, have long been a favourite case study for biologists. The search “bacterial flagellum” at PubMed generates 2 684 hits as of today, including 274 review articles. Without getting too deep into the literature, the flagellum is a complex, self-assembling piece of molecular machinery. Many of its proteins seem to be derived from other cellular machines, most notably secretory systems.

Parts of the science blogosphere are currently alight with debate regarding a recent PNAS paper which suggests that the 24 core flagellar proteins are derived from a single ancestral gene. Most of the debate is taking place over at The Panda’s Thumb. It’s been interesting to follow the progression from an opinion piece, through a rather dubious protein structure analysis (an area in which the main protagonist is clearly not an expert) to an attempt to replicate the findings.

That last point is important. There’s been a lot of comment about “trial by blogosphere” in this case. I have no problem with informed scientific debate on the Web – but let’s behave like scientists. If you take issue with someone’s data and findings, your very first action should be to analyse their data; either show that their analysis is wrong or your analysis is better. Otherwise it’s easy to come across as just another angry, opinionated blogger screaming into the void.

4 thoughts on “The great flagellum debate

  1. Yeah — although I’m not a big fan of BLASTology over serious phylogenetics, this whole critique of “well, the proteins can’t be homologous; their structures don’t look the same” is on par with “whales can’t be related to cows; they don’t look the same”. If molecular evolution has taught us anything, it is to reject the assumption that physical resemblance or the lack of it is conclusive phylogenetic evidence.

    Still, I’m schedule to present in a couple of weeks at our local journal club, and I might take Howard’s paper and the blogosphere’s response as my topic. It’ll be interesting to see what the response of non-blogging scientists will be.

  2. Be sure to let us know how it’s received.

    A lot of the debate is centred on inappropriate use of BLAST. It does seem rather simplistic in this case and it will be interesting to see what comes from the reanalysis that some people are now doing.

  3. You say that Nick clearly is not an expert on protein structures and imply that his argument suffered because of it. Well according to one structural biologist whose opinion I respect, his posting of protein structures clearly demonstrated that the two proteins were not homologous. He couldn’t understand why anybody didn’t understand this.

  4. his posting of protein structures clearly demonstrated that the two proteins were not homologous

    I wasn’t impressed. The post showed two structures at different scales, different rotations and using different graphical representations, then said “they look different”. As Jonathan pointed out, “eyeballing” is not a good argument. Now, if there’d been an attempt to superimpose the structures and some statistical analysis such as the RMSD, I’d have been impressed.

Comments are closed.