IEDG2008: Model systems are dead, long live model systems


I’ve discovered a couple of important things at this meeting.

One, late night sessions at west coast meetings are deadly for any of us coming from more eastern time zones. At least the morning sessions are low stress.

Two, I haven’t heard one Drosophila talk yet, and the message is clear: we’re now in the stage of evo-devo in which everyone is diversifying and chasing down a wide array of species. There was a bit of model-system bashing, but at the same time, everyone is acknowledging the crucial role of those traditional, but weird and derived, lab critters in providing a point of comparison and being the source of many of the tools being used to explore phylogeny now. I thought, though, that the smartest comment of the evening was that now everything is a model system.

I’ve got some dense piles of notes on the evening session, but I’m going to give you the short version of everything, with an emphasis on the novel twists.

Michael Akam talked about segmentation genes, which every developmental zoologist now knows inside and out — trust me, this is a familiar topic with over 25 years of detailed research … in Drosophila. Akam made the point that now it’s looking clear that three of the major segmented phyla, the arthropods, annelids, and chordates, may be using related genes to accomplish segmentation, but they seem to be using different mechanisms — so he considers the question of whether segmentation in these three is homologous is still an open question. He also discussed recent work on the centipede Strigamia (definitely not a lab animal: they can’t breed them in the lab yet, so all the work is done by collecting embryos in the field, in Scotland). They have a dynamic pattern of segment addition that is very different from what you find in flies, and more similar in some ways to chodate segmentation.

Chelsea Specht talked about floral evolution in the Zingiberales. I’m an animal guy, so even the most basic stuff in this talk was entirely new to me. I know the general rules of the spatial development of in the fruit fly of the plant world, Arabidopsis, and she gave us a bit of context there, reminding us of the concentric development of sepals, petals, stamens, and carpels. The Zingiberales are a large and diverse group of plants that includes bananas and ginger, and one characteristic is an extravagant modification of the canonical pattern, with extra stamens, a loss of select stamens, and a fusion of stamens to form a novel structure, the labellum, which in these plants functionally replaces the petals. So of course they’re looking into the genes involved in the patterns, which turn out to be the familiar Arabidopsis genes redeployed in new patterns.

Paul Sereno had a talk that took a very different tack, and was unfortunately giving it at the equivalent of 11:00pm Minnesota time, so I’m sorry to say I didn’t follow it carefully. He was discussing the analysis of morphology, and was advocating the development of tools and techniques to compare data sets in addition to the usual output, phylogenetic trees. He was making the case that a lot of morphological studies are actually very poor (a creationist in the audience would have loved it, largely because he wouldn’t have understood the context) because the input data sets of different studies are not comparable.

And now I have to get back to work and listen to the next set of talks.

Comments

  1. amph says

    …late night sessions at west coast meetings are deadly for any of us coming from more eastern time zones.

    Perhaps now you get an idea what it is to fly in to the US from Europe and go straight to the evening session the same day. Fortunately, I don’t snore when I am sleeping.

  2. says

    Yes, but they’re bound to be using the “model organisms” as a standard against which the others are (in part) understood. There’s certainly been a lot of value in studying model organisms.

    But yes, breaking out from the model organisms has always been a distant goal, and now it looks to be happening.

    Glen D
    http://tinyurl.com/2kxyc7

  3. frog says

    Your comment on Specht talk reminded me of a question I’ve always had: why aren’t there more topologists involved in evo-devo? It seems that development is almost entirely a topological problem (with a little geometry), so much of it could be formalized by some guys with Ph.D’s in topology.

    But I haven’t seen any work in that since Bateson (the dad) was pushed aside at the beginning of the 20th by the creation of neo-Darwinism.

  4. Josh says

    Paul’s right. It’s very difficult to use another person’s dataset, even if she/he works on the same organisms you do. The metrics are almost never the same. And frickin’ cladistics handles little unimportant things like variation so poorly that…well I don’t even go there.

    I’m a little surprised to see Sereno advocating studies along these lines though.

  5. says

    Apparently, Scotland is crawling with the wee many-legged beasties. It’s like centipede city everywhere, crawling up kilts, and clogging bagpipes.

  6. Mu says

    What’s an evening session? That’s why I love to be in Material Science, evenings are reserved for hospitality suites.

  7. caynazzo says

    I doubt whether you should read into the lack of fruitfly talks at this conference as indicative of a new trend. Yesterday, Mohamed Noor at Duke University gave a very well-attended talk at the NIH titled, Evolution and the Concept of Species. His model was the fruitfly.

  8. caynazzo says

    Oh and, Dr. Mohamed inspired nervous laughter when he took a moment to poke fun at the profound ignorance of Intelligent Design and a grinning Kurt Cameron popped up on the screen.

  9. says

    Apparently, Scotland is crawling with the wee many-legged beasties. It’s like centipede city everywhere, crawling up kilts, and clogging bagpipes.

    It’s all those damn haggises, you know.

    Doesn’t matter how many you choke down, there’s always more of them, crawling into food, totally corrupting any refined taste buds.

    Glen D
    http://tinyurl.com/2kxyc7

  10. Lord Zero says

    A youtube videoupload of the meeting would be nice,
    for more people to enjoy… im kinda envious, sounds
    so interesting. I liked to be there too, but im too busy…

  11. David Marjanović, OM says

    It’s very difficult to use another person’s dataset, even if she/he works on the same organisms you do. The metrics are almost never the same.

    Well, often there are no metrics at all. Just today I came across two characters in a published matrix (appendix to a peer-reviewed paper — the matrix itself was evidently not peer-reviewed…) that talk about whether the suture between two skull bones is “strongly interdigitated” or not. No attempt to define “strongly” is made, and the whole spectrum of imaginable morphologies between straight and “strongly interdigitated” occurs in the taxa that are in the matrix. It is flat-out impossible to replicate the authors’ scoring. This is not science. Fortunately, it’s not the best that can be done. If you know what you’re scoring, you can quantitatively describe it and publish that description. The problem is not one of principle.

    And frickin’ cladistics handles little unimportant things like variation so poorly that…well I don’t even go there.

    Do you mean intraspecific variation? Read a few Wiens papers in this case. I haven’t done that, but he cites himself as having developed impressive weighting schemes that seem to deal with this very nicely.

  12. travc says

    Thanks, I love reading this sort of stuff.

    Though, the completely unsurprising result that the same sets of genes are involved in different patterns always annoys me when it is reported as if it isn’t exactly what we expect.

  13. DaveH says

    Glen D @#15 “It’s all those damn haggises, you know.

    Doesn’t matter how many you choke down, there’s always more of them”

    Actually, the centipedes keep the numbers down by preying on haggis eggs. Too many centipedes taken from the wild might disrupt the balance, then you’ll see a plague of haggi. It’ll be like Exodus all over again, all because of the godless Evo-devo researchers!

  14. Josh says

    No attempt to define “strongly” is made, and the whole spectrum of imaginable morphologies between straight and “strongly interdigitated” occurs in the taxa that are in the matrix. It is flat-out impossible to replicate the authors’ scoring.

    Oh, shit. Let’s not even go there. One of my favorites has always been “The brevis shelf is wide.”

    Do you mean intraspecific variation?

    Good catch. Yes, that’s what I meant. I should have proofed the comment more.

  15. says

    This blogging from conferences is a wonderful idea. I think Tommaso Dorigo over at Quantum Diaries Survivor may have set the standard at a recent physics conference in Albuquerque. (There’s a whole string of comments — you can start here:
    http://dorigo.wordpress.com/2008/05/23/liam-mc-allister-inflation-in-string-theory/)

    Which brings up a question: Hey: Who is coordinating the blog fest from Evolution 2008? Can somebody set it up so there are a hundred posts on substantive things we can use to fight the Texas State Board of Education and Commissar of Miseducation McLeroy? Stuff teachers can use in the classrooms?

  16. randy says

    I was going to go to the IEDG meeting, but found some money to go to the Euro Evo-devo meeting in Belgium. Looking forward to good chocolate and bread.