Wednesday, July 21, 2010

Digging into the GER details

MSM and GER calculations are a powerful and objective way to determine the level of fit between any proposed cladogram and the fossil record.
We saw earlier the GER values obtained by Wills et al (2008) for 19 dinosaur groups. You will recall that the Paraves values were the lowest on the list.
To understand the deep significance of the Wills et al study we need to look at the 19 groups (taxa) that were included.
The 19 number is made up of 17 dinosaur groups such as Sauropoda, Hadrosauridae etc. These have high GER scores.
The other two are bird related groups - the Turner et al (2007) Paraves study and the Senter et al (2007) Paraves + Oviraptorosauria study.
The GER for the Turner et al (2007) Paraves was .5579. The GER for the Senter et al (2007) Paraves + Oviraptorosauria was .6780.
These are exceptionally low values indicating low congruence.


Three things to note:
1. When analyzing a strictly dinosaur group (such as Hadrosauridae, Sauropoda etc) the GER value is "extremely high" according to the authors. This is not surprising. In those cases we are not even analyzing the dino-to-bird theory.
2. However when we analyze a bird group (such as Paraves) the congruence is "indistinguishable from random".
3. The best way to analyze the dino-to-bird theory would be to calculate the congruence for coelurosaurs or theropoda. In other words a grouping that includes dinosaurs and birds. Unfortunately the Wills et al study does not include any study like that. But the Pol and Norell (2006) study does. (I will elaborate on that in the next post).

6 comments:

  1. Prosauropoda 23 0.6203
    Stegosauria 10 0.5789
    Sauropodomorpha (basal) 34 0.5674

    How about not just selecting the figures that you think support your argument?

    ReplyDelete
  2. The GER values are listed in order from Hadrosauridae at the top with 0.9561 to Paraves at the bottom with 0.5579.
    The bottom 5 are:
    Senter (2007) Paraves + Oviraptorosauria 0.6780
    Prosauropoda 0.6203
    Stegosauria 0.5789
    Sauropodomorpha (basal) 0.5674
    Paraves 0.5579.

    You stopped just short of including the Paraves + Oviraptorosauria in your list.
    How about not just selecting the figures that you think support your argument?
    The fact remains that Paraves and Paraves + Oviraptorosauria are at the bottom of the list.
    The hypothesized cladograms for those categories are not congruent with the fossil record.
    Did you notice the MSM* calcs for theropoda in the other post. What are your thoughts about that?

    ReplyDelete
  3. "You stopped just short of including the Paraves + Oviraptorosauria in your list. " Only because I was just posting the bits you dishonestly ignored. How about addressing the point - your figure for Paraves is not that anomalous in the context of the rest of the data even if it is at the bottom - especially since we get a similar figure for stegosaurs and protosauropods

    ReplyDelete
  4. My first observation is your use of the word "dishonestly".
    Your immediate inclination is to attribute dishonesty to someone that you disagree with. You try to personalize the discussion and attribute negative qualities to those you disagree with. I am very familiar with that tactic. It is a form of ad hominem attack. It means that you do not feel your own point of view is strong enough to stand on its own and you need to distract from the actual subject at hand by attacking the person you disagree with. And often enough that distracting tactic works when the other person falls for it. Try that tactic on others, it won't fly here
    My second observation is that I have posted 150 entries outlining an entire theory. Are you even aware of the major points I am making in this site?
    My next observation is that you seem not to be interested in the fact that the Paraves GER number is "indistinguishable from random". May I suggest that you let that sink in a little bit? It means that we have no reason whatsoever to believe that it reflects the actual phylogenetic relationships between the creatures it covers. The hypothesized Paraves relationships are no more accurate than if we had thrown darts at a wall.

    Moving to the actual point you are making. The table is sorted by GER value - with the highest at the top and the lowest at the bottom. Most dinosaur groups have numbers close to 1.0. In fact the authors of the study make that exact point:

    "Previous piecemeal investigations of a modest number of dinosaur clades have demonstrated remarkable congruence between phylogeny and stratigraphy (Brochu and Norell [2000] and Rauhut [2003] for theropods; Wilson [2002] for sauropods; Pol and Norell [2006] for basal sauropodomorphs). However, it is unclear whether this is simply a sampling effect given the trees so far investigated or whether exceptional congruence is a feature of dinosaur cladograms more generally. Here we provide the first systematic investigation of congruence between the dinosaur stratigraphical record and phylogeny on the basis of a sample of recently published dinosaur cladograms. By applying modified congruence indices to these data, we demonstrate that in half of our sampled cases, the fit of cladograms to the fossil record is as good as it could possibly be, whereas in half of the remaining cases, the fit is within 10% of the theoretical maximum. We also adapt a procedure implemented by Pol and Norell (2006) for incorporating uncertainty in the dating of the earliest occurring fossils. Even allowing for dating errors and artificially selecting the least congruent combinations of possible first occurrence dates, a quarter of dinosaur trees were still perfectly congruent, with another quarter within 10% of this optimum."

    The authors themselves draw our attention to how good the fit is for most of the dinosaur groups. My point is that since the fit is good for most of the dinosaur groups then the low Paraves number is all the more credible and significant. And even if we had no other numbers to compare it with we still know that it is "indistinguishable from random".

    Now that I have answered your question could you please answer mine?
    Please provide us with a reference for an MSM* value of 0.31 - 0.40. How should we evaluate that number? This is the most significant number of all, because it is for a cladogram that covers both actual dinosaurs (like tyrannosaurs and other non-maniraptoran theropods) and maniraptors.
    Let's focus on the heart of the matter.

    I know full well from lots of experience that people will hold onto any tiny, tiny point and argue it till the cows come home. I will not be falling for that here.
    I welcome your constructive contribution but this is all I intend to say on this particular issue.

    ReplyDelete
  5. I'm afraid your accusation of "ad hom" won't wash - you are being dishonest in your selective use of data, very much like the way creationists use quote mines in fact. There is nothing "ad hom" about accusing a person of misrepresentation if that is what they are doing. Lets look at this issue of the "GER" data for Paraves being "indistinguishable from random" I think if we post the quote from Will's et al in full it becomes clearer and again shows you have taken their words out of context:

    "GER∗ values for variable dates had a mean of 0.981, with 10 of 19 data sets having a mean (and minimum) of 0.999 (the highest value possible for 1,000 randomizations). Even the worst possible datings yielded a mean of 0.956, with just two data sets below 0.950. The interpretation of the GER∗ is more straightforward than that of the
    GER or GERt. GER∗ values above 0.500 indicate greater congruence than the median of the null distribution, and values above 0.975 indicate significantly greater congruence than the null distribution (with P ≤ 0.050, not correcting for multiple tests). For the static first occurrence dates, 16 of 17 data sets were significantly congruent, with just two being indistinguishable from random."

    Note first that they are referring to their adjusted GER* here, not to GER as you have erroneously stated. This is not an insignificant point, the main aim of the paper is to "propose modifications to an existing index of congruence (the gap excess ratio; GER), designed to remove a bias in the range of values possible with trees of different shapes. We also propose a more informative index of congruence - GER∗ - that takes account of the underlying distribution of sums of ghost ranges possible when permuting stratigraphic range data across the tree". i.e. the paper contains a critique of the use of GER and proposes a more robust measure, GER*.

    Secondly, and this is pertinent to my earlier point about how misleading it is to focus on paraves and leave out the other groups like stegosauria that also have comparatively low GER and GER* scores, the 2 groups they are referring to as having scores "indistinguishable from random" are in fact paraves (GER* 0.911) and stegosauria (GER* 0.781).

    It's important here to also place this in the context of the following paragraph. Remember that for GER* any score above 0.500 indicates “greater congruence than the median of the null distribution, and values above 0.975 indicate significantly greater congruence than the null distribution” – the use of the term “indistinguishable from random” is referring here to statistical significance – scores above 0.975 are considered to be statistically significant. The following paragraph places this data for all dinosaurs into context:

    “Preliminary reanalysis of the 1,000 animal and plant “static” data sets utilized by Benton et al. (2000) and Wills (2007) yielded an average GER∗ of 0.688, with only 34% attaining a GER∗ > 0.975. This strongly suggests that the congruence of dinosaur phylogenies is better than that for a large sample of data sets across a range of other taxa.”

    I think this speaks for itself. If you are going to try making use of papers like this in future you would be wise to quote in full rather than selectively picking and choosing what to quote as per your biases.

    ReplyDelete
  6. As I said:
    I know full well from lots of experience that people will hold onto any tiny, tiny point and argue it till the cows come home. I will not be falling for that here.
    I welcome your constructive contribution but this is all I intend to say on this particular issue.

    ReplyDelete