Saturday, July 07, 2012

Quantitative Trait Evolution

I'm at Evolution Ottawa 2012 and this afternoon I went to a session on New phylogenetic methods for quantitative trait variation.

I attended two talks.

Introduction and asymptotic behavior of trait evolution models under drift and stabilizing selection by Cecile Ane.

Placing fossils on molecular phylogenise with Brownian motion or Ornstein-Uhlenbeck models of continuous trait evolution by Joe Felsenstein.

It was interesting to be exposed to this kind of theoretical population genetics but I really didn't understand a single thing.

There were more than 200 people in the audience listening with rapt attention. They're a lot smarter than me. It was embarrassing.

There's one thing Joe said that I understood, and it was very impressive. He referred to a paper "published by me." Nowadays, almost everyone would say, incorrectly, "published by myself."


That's a photo of one of the speakers. Can you guess which one?

15 comments :

  1. What I want to know is why I look there like someone has asked a stupid question, and I am making a face -- since nothing like that happened.

    ReplyDelete
  2. This is the problem with the subject of evolutionary biology. Biochemists like Larry don't understand population genetics, and population geneticists like Joe don't understand molecular/cellular biology. How can you have a "synthesis" when there are so many disparate fields and everyone talks past one another? This is, I believe, due to academic "over-specialization" - all these experts occupy their own intellectual niche and so can no longer adapt in response to new findings. The next generation of scientists has to be trained as generally as possible and participate in cross-disciplinary research.

    ReplyDelete
  3. "Biochemists like Larry don't understand population genetics, and population geneticists like Joe don't understand molecular/cellular biology."

    And creationists don't understand any of it...

    ReplyDelete
    Replies
    1. There are some creationists in the population genetics business - John Sanford and Walter Remine among them. There are many more in the biochemistry business. Someone like Sanford is a truly cross-disciple scientist. He understands genetics and mathematical biology very well indeed.

      Delete
    2. John Sanford and Walter ReMine show astounding incompetence when they presume to understand more than their disciplines, yet make kindergarten-level mistakes. Neither of them understands population genetics. If those are your heroes, no wonder you so often make such displays of imbecility. Or should I say IDiocy?

      Delete
    3. Sanford is an IDiot and I am not surprised that you, Atheistoclast, would assert your latest breathtaking inanity. Moreover, I am well aware how poorly misinformed John Sanford seems to be not only with regards to population genetics, but indeed, with all of current evolutionary biology.

      John Kwok

      Delete
    4. I agree with NegEnt and Anonymous. ReMine claims that the 1667 fixed, beneficial mutations "allowed" by Haldane's model cannot explain human evolution from an apelike ancestor, yet cannot say what traits this ancestor had such that the 1667 he allows is too few! When asked to name some of these traits, he invariably refers to things like music appreciation, poetry writing, etc. - too ignorant of anatomy and physiology to realize that these things are all housed in and are thus 'features' of a single 'trait' - the neocortex!

      Then there is Sanford, who implies in his book that to get a 'new gene' via Haldane's model you have to have a succession of mutations to make it de novo. You'd think he and his pal ReMine could at least get together to coordinate their stories.

      That creationists think either of them make valid points demonstrates the naivete of the typical creationist.

      Delete
  4. How about the bit near the end when Joe talked about recognizing ancestors in the fossil record? That was mercifully free of mathematics. But Joe, you didn't consider another source of ambiguity: the ability to recognize species from preserved material. To use the cliche examples, how many species of australopithecines are there in the known sample? We can't really be sure of even that, even in a well-sampled (we think -- another can of worms) and depauperate group. How many species is H. erectus, really? So who's your ancestor now?

    ReplyDelete
    Replies
    1. A serious issue. I thought I was grappling with enough hairy issues as is. I suspect that a population perspective on australopithecines, etc. would be enormously helpful. The current situation seems to bias paleoanthropologists towards splitting, when some of those multiple "species" are probably just geographic variations. So you have raised a great issue but one for another day.

      PS John, look for Larry at the meeting. He's tall, white hair and white mustache, and so far wearing no jacket. I'll point you and him out to each other if I can.

      Delete
  5. John,

    I didn't realize that you were here. I don't have your email address. Please get in touch with me.

    ReplyDelete
    Replies
    1. Larry,

      Sorry I missed you. The meeting's internet connection blew up at just the wrong moment, and now I'm already home. Next time, then. I never managed to try a beavertail either. Which is the greater loss?

      Delete
  6. "They're a lot smarter than me."

    I doubt that's true.
    I bet they understand natural selection better than you (and me) though.

    ReplyDelete
  7. These topics are Fundamental to biology. In an ideal world, all biology students would get some exposure to them at senior undergraduate level, along with suitable preparation (i.e. relevant introduction to statistical modelling) for this at junior undergraduate level. Not teaching statistical modelling (as opposed to recipe-based statistics) to junior biology students is what leads to the fact that very few biologists understand the fundamentals of their field.

    (As a side effect, this also leads to one of Larry's pet peeves - biology being invaded by physicists, computer scientists and engineers, who _do_ get the relavant statistical training but lack the biological background.)

    ReplyDelete
  8. They're a lot smarter than me.

    Shouldn't that be "...a lot smarter than I"?

    Watch the copula. ;)

    ReplyDelete