Thursday, October 28, 2010

Giving Both Barrels: A Two-Part Response To Professor Larry Moran, Part 1

    I had the guts to make several comments on Professor Larry Moran’s blog, Sandwalk; he has responded to one of these comments with two blog posts (here and here). And now, I have the chutzpah to respond to Professor Larry Moran (to get the gist of what’s going on see my comments on Sandwalk) on BioTalk.

The comment to which Prof. Moran responded to was pretty straightforward. Namely, I argued that “intelligent design proponents need only demonstrate that the odds of a particular biochemical system evolving are 10^-40 or less in order for intelligent design to be a more adequate explanation for the origin of such a biochemical system. This is because there have been no more than 10^39 bacterial cells in the history of life on earth.”
       
    Prof. Moran contends that the above argument cannot possibly hold much water, and he attempts to substantiate that assertion by bringing up the following argument:
That the odds of a specified sperm uniting with an egg is deplorably low, and as a result, through various calculations he does, he concludes that the probability of his four great-grandparents being born is lower than 10^-64 – a probability significantly lower than my stated 10^-40.
     
        Unfortunately by arguing thusly our dear professor is opening a can of worms for the entire common descent framework in that he is tacitly “refuting” the evidence for common descent. One of the oft-cited evidences for common descent are the shared insertion sites of endogenous retroviruses between humans and chimpanzees. However, if we are to accept Prof. Moran’s dictum that something having a probability of occurring of 10^-64 is in fact perfectly plausible, then we must accept that it is perfectly plausible that the shared ERV insertion sites between humans and chimpanzees may be the result of mere coincidence. Prof. Moran is trying to have it both ways but that is a of course a  blatant contradiction.
       
   Moreover, I would like to ask Prof. Moran to please explain to me in terms of probability, why precisely polyvalent antimalarial treatments are more effective than standard treatments? It seems as if Prof. Moran is wishing to go against the peer-reviewed literature, where we see that White (2004) says,
"If two drugs are used with different modes of action, and therefore different resistance mechanisms, then the per-parasite probability of developing resistance to both drugs is the product of their individual per-parasite probabilities. This is particularly powerful in malaria, because there are only about 1017 malaria parasites in the entire world. For example, if the per-parasite probabilities of developing resistance to drug A and drug B are both 1 in 1012, then a simultaneously resistant mutant will arise spontaneously every 1 in 1024 parasites. As there is a cumulative total of less than 1020 malaria parasites in existence in one year, such a simultaneously resistant parasite would arise spontaneously roughly once every 10,000 years — provided the drugs always confronted the parasites in combination. Thus the lower the de novo per-parasite probability of developing resistance, the greater the delay in the emergence of resistance.”


Now, to support my assertion that it is implausible to expect Darwinian processes to produce a biochemical system having a probability of origin less than 10^40, we must merely use a little logic.
If the per-E.coli probability of evolving biochemical system A is 10^40, then a mutant E. coli with this biochemical system will arise every 1 in 10^40 cells. As there is a cumulative total of about 10^30 bacterial cells in existence in one year [Whitman et al. 1998], such a mutant E. coli would arise roughly once every 3.55 billion years.
      Sound familiar? I’m using the exact same math that White 2004 is using, yet amusingly I don’t see Prof. Moran talking about how the referee(s) of White’s paper messed up in their review of his paper; nor for that matter do I see him criticizing N.J. White’s work; nor for that matter have I seen a letter from Prof. Moran to the Journal of Clinical Investigation correcting White’s work. I wonder why that could be? Oh right, I forgot, I’m advocating intelligent design so only I can be wrong when I use the same reasoning a peer-reviewed paper uses. Neat, isn’t it?


References:

1. White, Nicholas J. Antimalarial drug resistance. J. Clin. Invest, 113(8):1084–1092 (2004). 

2. Whitman W.B, Coleman D.C, Wiebe W.J. Prokaryotes: the unseen majority Proc. Natl Acad. Sci. 95(12): 6578-6583 (1998). 


Addendum: I said that,
"If the per-E.coli probability of evolving biochemical system A is 10^40, then a mutant E. coli with this biochemical system will arise every 1 in 10^40 cells. As there is a cumulative total of about 10^30 bacterial cells in existence in one year [Whitman et al. 1998], such a mutant E. coli would arise roughly once every 3.55 billion years." 

I might add that " thus the lower the  per-E.coli probability of developing biochemical system A the greater the delay in the emergence of biochemical system A", a statement consistent with White's paper.









13 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. "Now, to support my assertion that it is implausible to expect Darwinian processes to produce a biochemical system having a probability of origin less than 10^40, we must merely use a little logic."


    First you must provide the evidence-supported rationale that:
    1. Any such system has such a probability and how that probability was determined.
    2. Your understanding of evolution is how actual evolutionary biologists understand it.

    From your posts on Larry's blog, it seems pretty clear that you ahve a very idiosyncratic view of what evolution is, you certainly have a rather, shall we say, 'unique' view on what 'darwinism' means.

    So I would like to see the actual work in calculating that a particular biochemical system has such a probability of 'origin' - such a calculation should employ rational, realistic, pertinent criteria and conditions (i.e., no 'it must all come together as-is all at once' deals).
    Keep in mind that starting from the end, that is, looking at a particualr biochemical system and working backwards is really an illogical and invalid way to look at this, since it implies that there was an impetus, a goal, for evolutionary processes to try to meet.

    "If the per-E.coli probability of evolving biochemical system A is 10^40, then a mutant E. coli with this biochemical system will arise every 1 in 10^40 cells. As there is a cumulative total of about 10^30 bacterial cells in existence in one year [Whitman et al. 1998], such a mutant E. coli would arise roughly once every 3.55 billion years."

    Providing, of course, that a pre-specified target is what evolution seeks to produce.

    Any evidence that this is the case?

    If not, then it seems to me that your entire line of reasoning is quite illogical - one massive strawman fallacy, cloaked in excessive verbosity.

    ReplyDelete
  3. Guts?

    Why guts?

    Larry's blog is not like Uncommondescent or ARN or any of the blogs or discussion boards run by YEC/ID types who either do not allow comments at all, or threaten those who do not agree with the party line with banning at the drop of a hat.

    But if I were you, I'd lay off the pedantery and try to write something of substance.

    ReplyDelete
  4. I really have no idea what you are trying to say about e. coli above, and I have some doubt that you do (have any idea, that is). If you want some scrupulously produced, quite wonderful experimental work on probabilities of occurrence of mutations in e. coli (including changes that necessitate two mutations) and regarding the time frames involved, Richard Lenski's lab has produced some real figures for you, that contrast markedly with your utterly erroneous hand calculation above. (Not surprising, since you quite evidently don't understand either the biology or the math.)

    Oh, and if you like the White paper you quote from above, here's another short piece from it you may enjoy:

    "In the emergence and spread of resistance to antimalarial drugs, there are many parallels with antibiotic resistance (36, 37) — particularly antituberculous drug resistance, where, as for malaria, transferable resistance genes are not involved in the emergence of resistance."

    So I wouldn't go generalizing from White's paper to evolution-as-a-whole in the first place, let alone do so on the basis of your (mis)understanding of what he says.

    ReplyDelete
  5. 1. Any such system has such a probability and how that probability was determined.

    I do not have to demonstrate that there is such a system in existence – whether or not such a system with such a ‘probability of origin’ exists is off-topic. What we’re discussing here is my argument that if such a system is discovered (a system with a ‘probability of origin’ less than 10^-40) then intelligent design is a more adequate explanation for its origin.
    That said, the probability would be determined through an analysis of the biochemical system under consideration. For example, we might conduct directed enzyme mutagenesis experiments on that system to determine which residues are necessary to the function of that system, and the like. Mutation rates would also be taken into consideration, to name one of the factors in determining the ‘probability of origin.’

    “So I would like to see the actual work in calculating that a particular biochemical system has such a probability of 'origin' - such a calculation should employ rational, realistic, pertinent criteria and conditions (i.e., no 'it must all come together as-is all at once' deals).”

    This is still off-topic, and I daresay it may be bordering on the level of red-herring. Furthermore, regarding your “no ‘it must all come together as-is all at once’ deals” statement, I would like to ask you if the Q46R and the G85K substitutions (in the coat protein of the Pepper mild mottle virus) must occur simultaneously (i.e. occur in the same mutant organism), all at once, in order for the Pepper mild mottle virus to overcome the resistance gene Capsicum [see Genda et al. 2007]? Or is this not one of those “it must all come together as-is all at once (in order to function I suppose is what you meant, but I’m not quoting you on that in case it’s a quote-mine) ” deals? I’d be interested in what your thoughts are on this.

    “Keep in mind that starting from the end, that is, looking at a particualr biochemical system and working backwards is really an illogical and invalid way to look at this, since it implies that there was an impetus, a goal, for evolutionary processes to try to meet.”

    I could argue the exact same thing against the evidence for common descent from ERV probabilities. So I guess you don’t believe that ERVs actually are evidence for common descent, since it could have easily been a coincidence?


    “Providing, of course, that a pre-specified target is what evolution seeks to produce.”

    Take that in with the referee(s) of White’s paper, since I am using the exact same line of reasoning White’s paper uses. I expect you are currently writing a letter to the Journal of Clinical Investigation, correcting White’s logic, right?

    ReplyDelete
  6. References:

    Genda Yoshikatsu, Kanda Ayami, Hamada Hiroyuki, Sato Kyoko, Ohnishi Jun, Tsuda Shinya. Two Amino Acid Substitutions in the Coat Protein of Pepper mild mottle virus Are Responsible for Overcoming the L4 Gene-Mediated Resistance in Capsicum spp. Phytopathology, 97(7): 787-793 (2007).

    ReplyDelete
  7. “Guts? Why guts?”

    Apparently my sarcasm is far more subtle than I thought it was; thanks for confirming that.

    “Larry's blog is not like Uncommondescent or ARN or any of the blogs or discussion boards run by YEC/ID types who either do not allow comments at all, or threaten those who do not agree with the party line with banning at the drop of a hat.”

    That’s not my problem, since I am not deleting comments or threatening anyone with banning on this blog.

    ReplyDelete
  8. “If you want some scrupulously produced, quite wonderful experimental work on probabilities of occurrence of mutations in e. coli (including changes that necessitate two mutations) and regarding the time frames involved, Richard Lenski's lab has produced some real figures for you, that contrast markedly with your utterly erroneous hand calculation above.”

    This is a blatant red herring. We are not discussing whether or not there is in existence a biochemical system with a ‘probability of origing’ of 10^-40; nor are we discussing the odds of two mutations taking place in E. coli. You seem to have completely missed my point of citing White’s paper.

    What is my argument? It’s very simple:
    If intelligent design proponents found a biochemical system that had a probability of origin less than 10^-40, then intelligent design is a more adequate explanation. This is all I am arguing.
    To which Prof. Moran responds with his analogy of great-grandparents. I defend my position by demonstrating that I am using the exact same lines of reasoning as White’s paper. This is what you, Doppelganger, and Prof. Moran need to refute. You need to show me that I am not using the same logic as White’s paper is using. I have yet to see any of you do that. Instead what you do is bring up irrelevant arguments and statements that have no pertinence to the discussion at hand. So, instead of delving into the land of red-herrings, I suggest you demonstrate to me that my logic is not consistent with the logic White uses in his paper. Oh, and could I ask you to show me exactly where my “hand calculation” is erroneous?

    ReplyDelete
  9. The primary fallacy in your argument is what's sometimes known as the Texas sharpshooter fallacy after a Texan who fires at the side of a barn, then draws a target around the bullet hole and claims to be a great shot -- to avoid this fallacy, you need to draw the target (/choose the event you're calculating the probability of) independently of what actually happens.

    To illustrate why I accuse you of this fallacy, let me take a (highly oversimplified and unrealistic) example: suppose there are 10^10 possible biological systems, each of which has a probability of evolving of 10^-40 (per bacterium). Suppose that 10^39 bacteria have existed on earth. In that case, we should expect that about 10^9 of these low-probability biological systems will have evolved by now. While the probability of any particular system evolving is indeed low, it's pretty much inevitable that *some* systems will evolve.

    (I shound clarify just how unrealistic the numbers above are: we have no idea how many possible biological systems there are [just how many we've found], their probabilities aren't all the same, and we don't even have a good way to figure out what their probabilities are -- as Doppelganger said, it's much harder than you seem to realize.)

    The same fallacy appears in the examples that Larry gave: it's highly unlikely that any particular sperm will unite with an egg, but fairly likely that *some* sperm will. Similarly, it's highly unlikely that any particular bridge hand will be dealt, but inevitable that *some* hand will.

    Note that this is not the case of the ERV argument: shared insertions are predicted by common ancestry, independependent of their observation. Furthermore, there are multiple instances of shared ERV insertions; you might be able to write one off as a coincidence, but not all of them.

    It doesn't apply to drug-resistant malaria, either: in that case, the target is being drawn before the event (in fact, the entire point is to make the target as small as possible, so the malaria are less likely to hit it).

    By the way, William Dembski has done a fair bit of work trying to figure out how to avoid this kind of fallacy by adjusting the probability based on the number of other things that might've happened -- that's what his Complex Specified Information (see http://www.designinference.com/documents/2005.06.Specification.pdf) thingie is all about. His work has some problems in the math, logic, and such, but by far the biggest problem is that NOBODY EVER SEEMS TO USE IT. Seriously. If you'd framed your argument in terms of CSI (and done it right), you'd have wound up compensating for the sharpshooter fallacy, and not drawing fallacious conclusions. ('Course, that's because you wouldn't have been able to draw any conclusions at all... maybe that's why nobody ever seems to actually calculate CSI.)

    ReplyDelete
  10. "I defend my position by demonstrating that I am using the exact same lines of reasoning as White’s paper."

    Yes, and everyone is trying to tell you that doing so is where you are going off the tracks. White is talking about specified mechanisms -- particular mechanisms. Gordon is exactly right that you're committing the sharpshooter fallacy. You can see this in a quote from White's paper:

    "For example, if the per-parasite probabilities of developing resistance to drug A and drug B are both 1 in 10^12, then a simultaneously resistant mutant will arise spontaneously every 1 in 10^24 parasites."

    So White's talking about the odds of two particular concurrent mutations occurring in the same organism. Note that the odds of a bacterium developing one of the two mutations is just a bit worse than 10^11. Suppose there are not just 2 mutations that might lead to malarial resistance, but 1000. Now we have a 10^(-9) chance of developing some sort of malarial resistance. The particular mutation that confers malarial resistance is probably more like 10^(-12), but since this is just one of several acceptable answers, the odds are quite a bit better of getting at least one right answer.

    How many different ways are there to build a flagellum? Considering the redundancies in DNA protein encoding, there's definitely more than one, probably on the order of thousands or tens of thousands. I don't actually know how many ways there are to build a flagellum, but then, I'm not the one trying to fix a probability to the event. Do you know how many different ways there are to build a flagellum? Can you provide any evidence for your answer?

    ReplyDelete
  11. The other important point here is how you're defining "probability." If you're using the traditional definition of "ratio of number of trials with positive outcome to number of trials total" then you're SOL since we don't actually know how many trials there have been (or for that matter, how many de novo flagellum mutations have occurred in bacterial lines that have since gone extinct).

    ReplyDelete
  12. Gordon Davisson:
    The weakest link in your argument is that you overlooked the glaringly obvious fact that in the argument I present here, I do not detail how intelligent design proponents would determine the odds of a biochemical system arising via Darwinian processes. I merely argued that if, regardless of how, but if intelligent proponents determined that the odds of a biochemical system arising via Darwinian processes was less than 10^-40, then intelligent design is a more adequate explanation for the origin of that system.

    Concerning your example – an example that you admitted was unrealisitc – I respond that it’s just that. It’s unrealistic, and therefore can be easily countered with estimates obtained from observations. You said that “suppose there are 10^10 possible biological systems;” unfortunately, it’s just that: a supposition that is at odds with biological reality (if I understand what you’re trying to say). If we wish to talk about ‘possible biological systems,’ – systems, I might add, that must be functional in order for your argument to hold – I refer you to Oberai et al. 2006, where we see that there are no more than 10,000 possible functional protein folds. Since a protein is only (in general) functional if its primary structure represents the ability to fold, then I can confidently assert that your above supposition is irrelevant.

    You go on to say that “while the probability of any particular system evolving is indeed low, it's pretty much inevitable that *some* systems will evolve.” True, it is inevitable that some systems will evolve, and we see this in nature (a la nylonase et al.); however your statement only holds for protein functions that have a low level of specificity. For protein systems that require a minimum of a large amount of fairly specified amino acids in order to function, then it is not inevitable that that protein function will evolve.

    ReplyDelete
  13. “Note that this is not the case of the ERV argument: shared insertions are predicted by common ancestry, independependent of their observation.”

    It seems as if you misunderstood the point of my ERV argument. I countered Professor Moran’s argument of the odds of a particular sperm fusing with a particular egg by bringing this up. It was only meant to counter the professor’s absurd argument. I am fully aware that it is entirely implausible to expect the nested hierarchical pattern demonstrated in ERV insertion sites to be the result of coincidence.

    “It doesn't apply to drug-resistant malaria, either: in that case, the target is being drawn before the event (in fact, the entire point is to make the target as small as possible, so the malaria are less likely to hit it).”

    Again, this argument is testimony to your overlooking the simple fact that I did not specify how the probabilities would be calculated, but rather that assuming such probabilities were calculated to be 10^-40, then it is beyond the plausible grasp of Darwinian processes, and intelligent design is a more adequate explanation.

    References:

    Amit Oberai, Yungok Ihm, Sanguk Kim, and James U. Bowie. A limited universe of membrane protein families and folds. Protein Sci. 15(7): 1723–1734 (2006).

    ReplyDelete