Probabilities And Skepticism

I thought about including this in my previous thread, but it has grown so large that I suspect it would be lost in the abyss. If Skeptical Zone readers are interested I’ll write a series of these posts, in which I’ll develop a number of themes concerning why I abandoned evolutionary orthodoxy and became convinced that an inference to design is most reasonable.

As most of you know, I am a classical musician. All great musical compositions have a theme, and the theme of this site is “think it possible that you may be mistaken.” With that theme in mind, might I suggest some skepticism concerning probabilities?

One doesn’t need precise numbers to recognize when proposed probabilities are way of whack. When I was growing up and learning mathematics my dad (a professor of chemical physics) admonished me to always check my calculations to see if they made sense on the surface (in my engineering department we call this “using the beverage out the nose” test). If I punch 87 x 53 into my calculator and get 46,481 I immediately know something is wrong (in this case I hit the 7 key twice by accident) even if I don’t know exactly what is wrong, because the result should be somewhere in the hundreds, not thousands. I don’t need to know exactly what the problem is in order to recognize that the result makes no sense.

I apply this logic to probabilities concerning evolutionary theory. We have some good empirical evidence that it took about 10^20 reproductive events for malaria to evolve chloroquine resistance. It could be that Lucy turned into Lizzie in 3.2 million years by stochastic Darwinian mechanisms filtered by natural selection, but I apply the beverage-out-the-nose test concerning the probabilities. Even given the most generous assumptions (a few hundred thousand generations with a few million individuals in each generation) the probabilistic Lucy–to-Lizzie resources don’t pass the smell test, in my view.

So, I ask Skeptical Zone readers: Is my skepticism unwarranted, and if so, why?

256 thoughts on “Probabilities And Skepticism

  1. RBH:
    Lizzie noted that

    We also need to have a pretty good idea of what is required to get from Lucy to Lizzie at the relevant level of analysis. Intuitive comparisons of gross morphology don’t do the trick.

    Yes, indeed. Specifically genetic variations involved in brain development. There are some potentially interesting candidates.

  2. Noam Ghish,

    ..192 human specific genes, arising in 5 million years. Let’s calculate the odds of that. The average human gene is 400 aa long. Let’s be generous and imagine that only 300 of those aa are absolutely necessary which is probably not true. So what’s 20^300, that’s 1 in 10^360? Well beyond Dembski’s probability bound of 1 in 10^150. There’s my proof, now let’s see yours.

    Once again, lets go back to the human specific TBC1D3 referenced in the paper you cited originally.
    http://stke.sciencemag.org/cgi/content/full/sigtrans;2/89/pe59
    TBC1D3 is a variant of the RabGTPase family TBC that appears to have evolved along the hominoid-lineage 35 million years ago by segmental duplication.
    http://www.pnas.org/content/100/5/2507.short
    Importantly, it’s human specific in that there has been a significant increase in *copy number*. Humans have 6 copies of the TBC1D3, chimps and orang-outangs have only two copies of this gene (go to PubMed protein, and do a BLAST search, then a similarity tree on BAK63846.1 TBC1 domain family member 3 [Pan troglodytes]).

    Here’s the first 60 amino acids of the Human vs chimp TBC1D3
    Query 1 MDVVEVAGSWWAQEREDIIMKYEKGHRAGLPEDKGPKPFGRYNNNVDHLGIVHETELPPL 60
    MDVVEVAGSWWAQEREDIIMKYEKGHRAGLPEDKGPKPF YNNNVDHLGIVHETELPPL
    Sbjct 1 MDVVEVAGSWWAQEREDIIMKYEKGHRAGLPEDKGPKPFRSYNNNVDHLGIVHETELPPL 60

    Your “probability” calculations are completely meaningless, we know gene duplications occur, we know the rates they occur at. You are carrying around at least a couple of gene duplications your parents didn’t have. Mutations that expand the copy numbers of the CYP450 genes are well known in medicine because they have important impacts on drug metabolism.

    Indeed if you look at the vast majority of “human specific” genes (and there was only 168 referenced in that paper), the story is the same one as for TBC1D3, where a human variant of an existing protein is expanded (in some cases not even expanded).
    http://www.broadinstitute.org/mammals/alpheus/data/human_specific_genes.txt

    Choosing a random gene from the above list, ENSG00000101446, this is the protein Kinase inhibitor SPINT3, a member of the KU superfamily. There are highly similar SPINT3 genes (better than 95% similarity) in chimps and orang-outangs, and very similar ones in other primates.

    Same for ENSG00000204919 (a proline rich protein of as yet unassigned function), and so on.

    There are three proteins that *are* unique to humans, which are frame shifts of genes that were present in the common ancestor of chimps and humans.
    http://genome.cshlp.org/content/19/10/1752.full

    Again, your “probability” calculations are irrelevant to the origin of these genes.

    Second, I would like to see the function of the gene when it only had 20 amino acids, then 21, then 22. I would like to see reasonable statistics for how such a gene could overcome the odds against its coming into existence by chance.

    Since no new gene originates this way, why would you like to see this?

    You conception of how new genes arise is a fantasy, which has no relationship to real biology.

  3. Petrushka:
    Butifnot,

    You and gpuccio should get together on this. He and Durston are the ones defining dFSCI as absolutely objective measures based on this “trivial” but objective definition of function.

    If you wish to define function in terms of utility for an organism, you are in the realm of natural selection.

    But how do you know whether or not a particular kind of stickiness would have utility in some system or another? When you invoke the magic word “design” what do u mean? What does the designer do in order to figure out what has utility and what does not? Describe the steps a designer takes when designing.

    I could describe the steps taken by Edison for any number of inventions. When you say design, do you mean magic, or something analogous to what human designers do?

    I think there is a fundamental error at the base of much evolutionary thinking – a ‘confusion’ between what something is and what it does. Utility – that’s the point, not that something is sticky, but how the stickiness is ‘used’. Stickiness = lump, stickiness as part of a system could be endlessly useful.

  4. damitall,

    Damitall you’re way off the page. Binding proteins is exactly nothing, let alone an immune “system”. Enzymes are what they are, to have function they have to be used.

  5. Noam Ghish: both writing and speaking have their pros and cons.
    I would more expect an honest seeker for truth to search for truth where he can find it, and to gather as much exposure for the search as possible, rather than to closet himself with a person while limiting the exchange to written rhetoric.

    Such assertions have no place in logical argument.To prove someone does not understand something you first have to show that they’re wrong on a certain issue.Since it’s very difficult to prove anything in philosophy beyond a shadow of a doubt, one might as well not even bother.

    That also applies to your claimed attempts to correct the misunderstandings of others. I have little interest in discussion with one who cannot separate his philosophy from empirical study.

Leave a Reply