Unfortunately, Professor Larry Moran has been caught quote-mining in his blog post God Plays Bridge. And he doesn’t save the best for last: he quotes-mine right at the start. In case he accuses me of quote-mining, I will quote the Prof’s quote-mine (this blog post is not intended to be a tongue-twister regardless of what you might suspect):
“The creationists tell us that anything in biology with a probability of 10^39 or less is impossible.”
I searched and searched the web and didn’t find any sites saying that a biochemical system with a probability of origin less than 10^39 is more adequately explained by an intelligence, so I can confidently conclude that Prof. Moran is referring to what I said, and only to what I said (I suspect that in lieu of being caught quote-mining Prof. Moran will attempt to claim that he wasn’t referring to what I said per se’; we shall await further developments). So, where does he quote-mine me?
For starters, Prof. Moran confuses “impossible” with “intelligent design is a more adequate explanation.” What I said precisely was this,
“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.”
Where did I say that if the odds of a particular biochemical system evolving is 10^-39 (or for that matter 10^-40) then it is impossible (to evolve I suppose is what Prof. Moran meant)? I did not say such a thing; I only said that in such a scenario intelligent design is a more adequate explanation for the origin of that system. Either I’m aging far sooner than I expected and Prof. Moran is seeing something I said that I can’t see or Prof. Moran is quote-mining. Take your pick.
This quote-mining business gets even more interesting when we see that Prof. Moran himself condemns quote-mining and has a lot to say about this dirty business (here, here, here, and here). Talk about double standards.
Anyway, Prof. Moran tries yet again to refute my argument regarding the probability of a Darwinian origin of biochemical systems, this time using a bridge game as an analogy (incidentally, I don’t play bridge). I believe I sufficiently responded to his analogies with a reference to a peer-reviewed paper, and once again I will remind our dear professor that his arguments contradict the very arguments advanced for common descent; namely, the argument from ERV insertion sites.
I wonder if Prof. Moran plays bridge? If he does, I wonder if that was when he thought up that lovely quote-mine?
This comment has been removed by the author.
ReplyDeleteIn case anyone suspects that I am removing Doppelganger's comments, then I would like to announce that I have not removed a single one of his comments and that he is the one removing his own comments (one really does wonder why).
ReplyDeleteActually, most biological systems evolve along single mutational paths. Thus for each evolutionary step there is 1/6000 chance of a particular mutation occurring per mutational event (for a 300AA protein). Which is highly likely to occur on evolutionary timescales.
ReplyDeleteWow you challenge professors! That takes a lot of audacity Living. I respect that! To bad I have no idea of what your arguing about.
ReplyDelete"Thus for each evolutionary step there is 1/6000 chance of a particular mutation occurring per mutational event (for a 300AA protein). Which is highly likely to occur on evolutionary timescales. "
ReplyDeleteI beg to differ on two accounts. Firstly your figure 1/6000 is grossly inaccurate. You forget to take into account the fact that in diploid organisms there are 175 nucleotide substitutions per generation [Nachman and Crowell 2000]. This greatly reduces the odds of a particular amino acid substitution taking place, given a large genome size.
Secondly, even assuming your figure is correct, this still presents a problem for a Darwinian origin of protein functions. If a protein function requires 24 very specific amino acid substitutions, then the odds of that protein function evolving is on the magnitude of 1/6000^24, or 10^-90, a probability obviously lower then the total number of bacterial cells that ever existed.
Thanks Chris ; )
ReplyDeleteOnce one realizes that professors are as human as you are, then it's a little easier to have the audacity to challenge their ideologies.
"I beg to differ on two accounts. Firstly your figure 1/6000 is grossly inaccurate. You forget to take into account the fact that in diploid organisms there are 175 nucleotide substitutions per generation [Nachman and Crowell 2000]. This greatly reduces the odds of a particular amino acid substitution taking place, given a large genome size."
ReplyDeleteWhich amount to about 4 amino acid substitutions per generation. Let's not quibble about these numbers. That means on the order of tens or hundreds of thousands of years there is a very likely chance of a particular single amino acid substitution in a particular gene to occur.
The key point is that most protein functions evolve through single mutational paths with monotonic fitness increases. Thus single mutations can be used to build up more complex adaptations. For a nice review on the subject, see http://www.ncbi.nlm.nih.gov/pubmed/19249235 .
"Which amount to about 4 amino acid substitutions per generation."
ReplyDeleteI have no idea where you pulled the number '4 amino acid substitutions per generation' out of. 175 nucleotide substitutions per generation would amount to ~58 aa substitutions per generation (175/3; 3 nucleotides make up one codon). Genome size and the rate of substitution therefore plays an important role in calculating rates of molecular evolution.
"The key point is that most protein functions evolve through single mutational paths with monotonic fitness increases. Thus single mutations can be used to build up more complex adaptations."
Single mutations can be used to build up more complex adaptations, but only some. There are some protein functions that require two specified mutations before any novel function is realized [see Genda et al. 2007].
Also, the paper you cited (I read the paper before you referred me to it) states that he majority of the potential amino acid sequences for a protein of moderate size (100aa) result in unfolded and therefore functionless protein structures. This has profound implications for the origin of protein folds.
Finally, I might add that it is an empirical observation of the biological world that there are many examples of novel protein functions evolving that require only one amino acid substitution, less examples of novel protein functions evolving that require two specific amino acid substitutions, even less that require three, and so on. This demonstrates that not all novel protein functions can evolve via a small number of phenotypically beneficial amino acid substitutions.
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).
"I have no idea where you pulled the number '4 amino acid substitutions per generation' out of. 175 nucleotide substitutions per generation would amount to ~58 aa substitutions per generation (175/3; 3 nucleotides make up one codon). Genome size and the rate of substitution therefore plays an important role in calculating rates of molecular evolution."
ReplyDeleteI'm taking into account % of the genome that codes, and silent mutations. This is for humans. I don't know why we're using human numbers, though. We are actually rather unimportant as far as evolution goes.
Here's something to consider: in a dense 5mL culture of E. coli, there is every possible single amino acid mutation, and many of the possible double amino acid mutations. In a single HIV-infected individual, there is every possible double amino acid mutation, and many triple mutations, for HIV proteins.
"Here's something to consider: in a dense 5mL culture of E. coli, there is every possible single amino acid mutation, and many of the possible double amino acid mutations."
ReplyDeleteThe problem is an exponential one. While the genome of an E. coli population may be mutated several times over, the odds of a novel protein function evolving that requires several dozen specified amino acid substitutions is next to nil. As I stated earlier, given your figure 1/6000 as the odds of a particular amino acid being substituted for another, then the odds of a novel function evolving that requires 24 specified amino acid substitutions are deplorably low, approx. 10^-90.
"In a single HIV-infected individual, there is every possible double amino acid mutation, and many triple mutations, for HIV proteins."
This is not really significant at all, considering that the size of the HIV genome is ~10,000 nucleotides, which would in effect be ~3333 codons. Even supposing that HIV reads the codons in its genome in an overlapping manner, this would result at most to 10,000 amino acids capable of being potentially mutated. Compared to the size of the genome of E. coli, this is hardly significant.
"the odds of a novel function evolving that requires 24 specified amino acid substitutions are deplorably low, approx. 10^-90."
ReplyDeleteCan I get a reference for this? I've never heard of any protein that requires such concerted mutations.
Also, keep in mind that viruses can have population sizes orders of magnitude higher than that of bacteria. That can explain how 2 simultaneous mutations can arise.
However, it is generally accepted that single mutant steps give rise to new protein functions.
“Can I get a reference for this? I've never heard of any protein that requires such concerted mutations.”
ReplyDeleteActually, a protein that requires only 11 specified amino acid residues working in concert in order to maintain that function would be beyond the plausible grasp of Darwinian evolution, by your own calculation (6000^11=10^41; a probability below the number of bacterial cells that ever existed). You therefore do not believe that there is in existence a protein or protein-based system that requires that level of specificity, correct? And if such a system were brought to the table, what would be the implication for Darwinian evolution?
“Also, keep in mind that viruses can have population sizes orders of magnitude higher than that of bacteria. That can explain how 2 simultaneous mutations can arise.”
I have never stated that a novel function that requires 2 simultaneous mutations is beyond the plausible reach of Darwinian evolution. I have however asserted that there are more novel functions evolving that require only one amino acid substitution than there are that require two amino acid substituions.
“However, it is generally accepted that single mutant steps give rise to new protein functions.”
Many times single amino acid substitutions do result in novel protein functions. However, to extrapolate this to the evolution of all protein systems is absurd at best. Firstly, there is an important but subtle difference between directed enzyme mutagenesis and the origin of novel protein functions in biological forms. There is a difference in requirements for the in vitro evolution of a novel function and the in vivo evolution of a novel function. In vivo novel protein functions must meet the stringent requirements of natural selection, which is not true for directed enzyme mutagenesis. So, my advice for you if you wish to cite any papers along these lines is to ensure that that paper is referring to in vivo evolution of novel protein functions.
That said, your statement that “it is generally accepted that single mutant steps give rise to new protein functions” holds little water if we are to assume this to refer to all protein systems. For example, Genda et al. (2007) speak of a novel protein function that requires two specific amino acid substitutions in order to be realized. But your claim holds even less water than you might suspect. If it is true that single mutant steps can give rise to new, highly specified protein functions, then that assertion must be reconciled with the data. Lack of function results in the FliG protein if one deletes segments (10aa each) GΔ1 to GΔ3, GΔ12, GΔ14, GΔ15, GΔ22 to GΔ23 and GΔ26 [Kihara et al. 2000]. May one then inquire how your assertion can be defended when one can patently see that, in FliG lacking the above segments, a single amino acid substitution will not result in any beneficial phenotype?
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).
Kihara May, Miller Gabriele U., and Macnab Robert M. Deletion Analysis of the Flagellar Switch Protein FliG of Salmonella. J Bacteriol., 182(11): 3022–3028 (2000).
"Actually, a protein that requires only 11 specified amino acid residues working in concert in order to maintain that function would be beyond the plausible grasp of Darwinian evolution, by your own calculation (6000^11=10^41; a probability below the number of bacterial cells that ever existed). You therefore do not believe that there is in existence a protein or protein-based system that requires that level of specificity, correct? And if such a system were brought to the table, what would be the implication for Darwinian evolution?"
ReplyDeleteI'd certainly be interested. Do you have such an example?
I'm not sure what you're trying to show with the papers you've mentioned. A single case of 2 concerted amino acid mutations in viruses does not disprove the general rule that evolution occurs through single mutational steps (which has been shown both in vitro and in vivo). I have no idea what you mean by your FliG deletion example.
Incidentally, have you read anything by Douglas Axe?
“I'd certainly be interested. Do you have such an example?”
ReplyDeleteThe bacterial flagellum is such an example. The absence of any of four critical residues (Arg90,Glu98, Pro173, Pro222) in the MotA protein in the bacterial flagellum results in abolished torque generation [Zhou and Blair 1997]. Residue Asp32 in MotB is also necessary for torque generation [Braun et al. 1999]. Three residues in FliG are critical for torque generation as well, namely Arg279, Asp286, and Asp287 [Lloyd and Blair 1997]. Finally, the deletion of any of some eight 10aa segments (segments 10, 18, 19, 24, 25, 28, 29 and 31) in FliG result in abolished torque generation [Kihara et al. 2000]. This means that flagellar torque generation requires all eight of these segments in order to function.
Here, I have identified a protein-based function – flagellar torque generation – that requires at least 8 very specific amino acid residues and another 8 very specific segments in order to function. This amounts to a grand total of some 16 very specific mutation events. The absence of any of these residues or amino acid segments results in abolished torque generation function in the bacterial flagellum. Thus, this would seem to more than fulfill the criteria of a protein-based function that requires at least 11 specified amino acid residues. True, in the above example, only (at least) 8 specified amino acids are required; however, another 8 very specific segments are also needed, and hence I think this example substantially fulfills the criteria. But I’m not stopping here. The function of torque generation in the flagellar motility system gets even more beyond the plausible reach of Darwinian processes. The absence of any of three specific 10aa segments in MotA and three specific segments in MotB results in no torque generation [Muramoto and Macnab 1998]. And that’s not all either. Site-directed mutagenesis at codon 261 and 268 in MotB suggest that these residues are critical for flagellar motility [Van Way et al. 2000].
In conclusion, the evolution of the torque generation function in the flagellar motility system would require some twenty-four (10 very specific amino acid residues and 12 very specific amino acid segments) specific mutation events in order for any function to be realized. As such, I can confidently assert that this function adequately meets the stated criteria (i.e. the example you requested).
“…the general rule that evolution occurs through single mutational steps (which has been shown both in vitro and in vivo).”
ReplyDeleteAgain, while this may be true in many cases of protein evolution, one cannot simply assume that this is true in every instance; and indeed it isn’t true in every case. Also, I am pretty sure that this ‘general rule’ you speak of is far more prevalent in vitro than in vivo.
“Incidentally, have you read anything by Douglas Axe?”
Yes I have read his papers, and I am currently reading his latest paper published in Biocomplexity, an intelligent design journal. After reading it I’ll probably have more to say on the problem of evolution navigating the vast size of potential protein sequence space. Oh, and incidentally, I didn’t intend this to be a research paper but it somewhat evolved into one (pun intended).
References:
ReplyDeleteZhou J, Blair DF. Residues of the cytoplasmic domain of MotA essential for torque generation in the bacterial flagellar motor. J. Mol. Biol, 273(2):428-39 (1997).
Braun TF, et al. Function of proline residues of MotA in torque generation by the flagellar motor of Escherichia coli. J Bacteriol. 1999;181:3542–3551
Lloyd SA, Blair DF. Charged residues of the rotor protein FliG essential for torque generation in the flagellar motor of Escherichia coli. J. Mol. Biol. 266(4):733-44 (1997).
Kihara May, Miller Gabriele U., and Macnab Robert M. Deletion Analysis of the Flagellar Switch Protein FliG of Salmonella. J Bacteriol., 182(11): 3022–3028 (2000).
Muramoto K, Macnab RM. Deletion analysis of MotA and MotB, components of the force-generating unit in the flagellar motor of Salmonella. Mol. Microbiol., 29(5):1191-202 (1998).
Van Way SM, Hosking ER, Braun TF, Manson MD. Mot protein assembly into the bacterial flagellum: a model based on mutational analysis of the motB gene. J. Mol. Biol., 297(1):7-24 (2000).
"Also, I am pretty sure that this ‘general rule’ you speak of is far more prevalent in vitro than in vivo."
ReplyDeleteAs far as I know, it holds both in vivo and in vitro. Do you have evidence that it doesn't?
I looked into the papers you cited. These studies look at mutations that destroy or weaken the function of flagellar proteins. However, they do not look into the evolution of function from an ancestral protein. It's not too difficult to destroy the function of a protein, but finding the mutational path it took from an ancestral protein is a tough problem.
What I was hoping to find is something like:
Darwinian evolution can follow only very few mutational paths to fitter proteins.
Weinreich DM, Delaney NF, Depristo MA, Hartl DL.
Science. 2006 Apr 7;312(5770):111-4.
In this study, they looked at all mutational paths that could be followed to evolve a beta lactamase to degrade a particular antibiotic, which took 5 mutations. There were indeed single-mutant paths that could be followed.
After a rejuvenating weekend, I will take the time to respond to your comments.
ReplyDelete“As far as I know, it holds both in vivo and in vitro. Do you have evidence that it doesn't?”
I am not, by any means, proposing that this ‘general rule’ is a rule that only applies in vitro; rather, I am arguing that this ‘general rule’ is far more prevalent in vitro than in in vivo. Do you have any citations that suggest that this ‘general rule’ is as prevalent in vivo as it is in vitro?
“However, they do not look into the evolution of function from an ancestral protein. It's not too difficult to destroy the function of a protein, but finding the mutational path it took from an ancestral protein is a tough problem.”
There are several problems with this argument. Firstly, you asked for an example of a protein function with the level of specificity of requiring 11 amino acid residues workin in concert, which I provided. Having said that, I will now explore the problems with your above argument.
Firstly, a brief ClustalX alignment will reveal that residues critical to torque generation in the flagellar motility system are conserved across species, with minor deviations consisting only of those residues that share similar charges, i.e. similar physicochemical properties. This implies that these residues were always critical, ever since all flagellar organisms shared their most recent common ancestor. Hence, the argument that there could be some torque generation through some less specified ancestral protein does not hold, as sequence alignments demonstrate that the critical residues I mentioned in my earlier comment were always essential.
Secondly, we can deduce from site-directed mutagenesis that the absence of any one critical residues in the flagellum result in abolished function. To argue that this is merely because it is “not too difficult to destroy the function of a protein” is absurd for the simple reason that no where in nature do we see a novel protein function evolving that requires say, a couple dozen very specific residues in order to maintain that function. If it were true that there are single-mutant paths that can result in a novel protein function that site-directed mutagenesis results in the conclusion that that function requires several dozen residues in order to function, then we would observe such protein functions evolving in vivo. But we don’t. This is perhaps the greatest problem with the above argument. I realize that I am probably not being that clear, so if you have any questions I will elaborate upon this.
“What I was hoping to find is something like…”
While you’re talking about beta-lactamase, you might want to cite just one peer-reviewed paper that demonstrates the real-time evolution of beta-lactamase. Also, I find it rather intriguing that you are asking me for a paper that looks at all the molecular mutantional paths for the function of torque generation, since if I presented such a paper, then that paper would support your position, and not mine. Indeed, the burden of proof is on you to cite relevant literature that would indicate that the function of torque generation in the flagellar motility system could indeed evolve through a single-mutant pathway.
"Indeed, the burden of proof is on you to cite relevant literature that would indicate that the function of torque generation in the flagellar motility system could indeed evolve through a single-mutant pathway."
ReplyDeleteLet's be clear here. The burden of proof is on you to show that there are unevolvable proteins. I'm just trying to help you think about your arguments.
For example, here's something you should know: EVERY protein has essential residues. Indeed, many biological systems are irreducibly complex. Can you tell me why they can still evolve?
(Also, I'd recommend against using Clustal for multiple sequence alignments.. it's the worst of the bunch)
“Let's be clear here. The burden of proof is on you to show that there are unevolvable proteins. I'm just trying to help you think about your arguments.”
ReplyDeleteI have demonstrated that there are unevolvable protein functions; torque generation in the bacterial flagellar motility system. To which you responded that we need to detail the mutational pathway that the ancestral protein took. I answer that there is no mutational pathway, and therefore providing such literature could only support your position.
“For example, here's something you should know: EVERY protein has essential residues. Indeed, many biological systems are irreducibly complex. Can you tell me why they can still evolve?”
Apparently you did not read my comments on Sandwalk (not that I expect anyone to). If you do take the time to read them, you will find that I said this:
“…all biological functions are irreducibly complex, and in that respect Behe (1996) is wrong in saying that IC systems cannot plausibly evolve via Darwinian processes. What is true is that biological systems that have high levels of IC are less and less likely to evolve the more IC they are.”
The reason that many IC systems can evolve is because those IC systems require fewer minimum components working in concert in order to function. Biological systems that have very high levels of IC, however, are beyond the grasp of Darwinian processes.
Regarding Clustal:
I agree, except that this wasn’t a formal experiment and Clustal is accurate enough for simple alignments ; )
You can't just claim that something can't evolve. You have to actually get in a lab and test it. It's tough, but I'm sure somebody with your knowledge can think up a few elegant experiments to do so.
ReplyDeleteAnd the reason that irreducibly complex systems can evolve is that their current function wasn't what was being selected for. For example, consider the citric acid cycle.
"You can't just claim that something can't evolve. You have to actually get in a lab and test it."
ReplyDeleteSite-directed mutagenesis can determine which residues are critical to a particular protein function, and thus it is a method to determine the 'evolvability' of a protein-based system.
"It's tough, but I'm sure somebody with your knowledge can think up a few elegant experiments to do so."
Thanks for the compliment. That aside, I am certain that there are in fact several potential experiments to test whether or not a protein-based system might evolve. However, site-directed mutagenesis, sequence alignments, and just basic observations in the biological world can give us confidence that a certain system is 'unevolvable.'
"And the reason that irreducibly complex systems can evolve is that their current function wasn't what was being selected for. For example, consider the citric acid cycle."
The analogy of the citric acid cycle is not that applicable to other protein-based systems, for the reason that the citric acid cycle is a cascade system, while say, torque generation in the bacterial flagellum is not. Actually, the reason that IC systems can evolve is because those IC systems do not represent high levels of IC.