More than a month ago now Jeffrey Tomkins published a DpSU on the ICR’s website called Study Shows Proteins Cannot Evolve. He also posted it on his own blog, Designed DNA, a couple of days later as he often does. Because it appeared the last day before exams I didn’t have much time to write anything more than a recommendation that you compare Tomkins’ article with a blogpost by fellow young Earth creationist Todd Wood on the same subject.
However, it appears that somebody else got in contact with Tomkins about it as on November 22 he published another post on his blog (have I mentioned that I still have a couple of posts from last month to do? I still have a couple of posts from last month to do), called Are Bigger Proteins More Favorable to Evolution? He opened:
I recently wrote an article discussing a recent research paper that showed how proteins contain non-negotiable sectors that are intolerant of amino acid changes and that when the other protein sectors that may tolerate such changes are altered, the changes often reduce quality of protein function. The 3-dimensional structure that proteins fold into, is also affected by the sequence of amino acids.
We’re going to need more than a little context. The original paper was The spatial architecture of protein function and adaptation in Nature – a pdf is available here. The bulk of the study involved modifying a relatively small protein and comparing how well it functioned compared to the original. The protein binds to a specific ligand (a small molecule), and the researchers experimentally determined what effect every single potential amino acid substitution (the swapping out of one, but only one, amino acid in the protein for another) on the protein’s ability to bind to its ligand.
There is a common misconception that most mutations to proteins are harmful: while technically true, it is more accurate to say that most mutations have very little effect either positive or negative, and that only a (sizeable) minority are seriously problematic to the functioning of the protein. This is shown in the figure reproduced above. Mutations with slight positive effects are in pale red, those with slight negative effects in pale blue, and those with more severe negative effects in darker shades of blue. For example, most mutations to #311 – the first in the chart, which is proline in the “wild-type” (see this table to convert the letters given into amino acid names) – seem to be slightly beneficial, while for #372 (wild-type: histidine) every mutation will produce a significantly less functional protein, though the degree varies.
The “spatial architecture” part of the research involved looking at which parts of the protein contained high concentrations of crucial amino acids. They found, curiously, that not all of the amino acids in the parts of the protein most closely involved in its activity are crucial (i.e. unable to be mutated freely). They also found that some of the amino acids that are crucial lie outside of this “active site.”
A third discovery was that a certain substitution created a protein that would bind both to the original ligand, and to a second, slightly different one. A second substitution removed the ability of this protein to bind to the original ligand, so that it now solely bound to the new one. While this isn’t directly related to Tomkins’ more recent post, it does demonstrate that protein evolution can still proceed in these circumstances despite his claims to the contrary. The reason that these mutations don’t show up as deep red points on the diagram is because in that experiment they only tested how well the protein functioned at its original job of binding to the original ligand. It is easy to observe how a mutation effects the usual function of a protein, but much harder to evaluate possible benefits of the mutation in relation to the billions of other potential functions for the protein in the context of the organism as a whole.
In short: some mutations are bad, most mutations do nothing of note. Tomkins writes:
In my article, I made the following comment…
“Imagine if this sort of experiment was done in more complex proteins that are hundreds of amino acids in length, or protein complexes that also include metal ions, carbohydrates, and ribo-nucleotides integrated into their structures.”
He implicitly argued – baselessly, in my opinion – that if you increase the complexity of the protein you would find that a greater proportion of the potential mutations would be significantly detrimental. Somebody called him out on this:
One critical point brought up by a reader of my article questioned the reasoning behind this statement which goes as follows…
“Wouldn’t a more complex protein with a longer amino acid sequence theoretically be MORE tolerant of single-base changes (outside of stop-code sequences)? I understand it is a leap to go from “tolerant of change” to an actual change in function or increase in complexity, but it seems your argument in the aforementioned article is the opposite: that increased protein size makes the protein less able to withstand change-of-base mutations.”
I don’t know if you could say that they should be more tolerant of changes either – I don’t know a huge amount about this topic and a lot of stuff is being asserted without proper evidence here. I will take as my null hypothesis that a larger protein is no more or less resilient than a small one, but if you know better I’d like to hear it.
Tomkins begins his response:
On the surface this question sounds reasonable. However, the logic that somehow by adding more amino acids, you can increase the odds of getting a favorable evolutionary outcome through random changes is a false line of logic in engineered systems. Larger non-structural proteins (enzymes, DNA-binding, etc) represent an incremental or commensurate increase in functional information because they have more complex features and perform more complex functions than smaller proteins. This is particularly true in multicellular organisms where the genome is contained in the nucleus and the cell system is considerably more complex than in a bacteria.
Yes, he is taking the assumption that proteins are “engineered systems” as a premise, and implicitly that they have the same degree of resilience as human engineered systems. His reasoning is therefore circular (he is using this to show design, after all), but that isn’t what is important here. The question is how the number of crucial amino acids scales as the total length of the protein increases (this is what the questioner was asking about, remember). If you triple the length of the protein will the number of crucial amino acids triple as well? Might it only double, quadruple, or not increase at all? Tomkins simply cannot tell us this information, and talking about complexity doesn’t help.
A good analogy would be the comparison of a wristwatch and a cell phone. The removal of a single electronic component from each system will still result in the failure of the whole system. The individual components (chips) in each system are more complex in the cell phone compared to the wristwatch, but just as critical. There is NOT more room for slop and error in the cell phone just because it is bigger or it’s components are bigger.
This is a poor analogy. Once again, the majority of the possible substitutions have little effect – hardly analogous to the “removal of a single electronic component from each system [causing] the failure of the whole system.” Because most mutations have little effect it is possible – indeed quite common – for slightly negative mutations to appear and stick without natural selection weeding them out. This is why there are so many potential positive mutations, as the protein is close to but not quite at its maximum efficiency (it doesn’t need to be). If a wristwatch was remotely analogous to our little protein then it would be a watch with a cracked screen and a frayed band, so battered that an appreciable proportion of all possible random hammer blows would actually fix some of its problems.
Is our cellphone equally damaged? We can’t tell, yet this is the important piece of information that we are trying to uncover. This is a bad analogy that tells us nothing.
The concept that larger proteins have more room for error or tolerate more slop is a fallacy. Indeed, the recent set of research papers regarding the sequencing of the human exome (protein codings regions of the genome) showed that variation in human proteins are not only rare, but in many cases are associated with heritable disease (1). Most of the genetic variation in the human genome is associated with non-coding DNA.
This is a red herring. Yes, some possible protein mutations will be selected against, but how many? The paper, of course, is “Evolution and Functional Impact of Rare Coding Variation from Deep Sequencing of Human Exomes,” which we have seen time and time again of late. And most of the genome is non-coding, after all, so exactly how much variation can we expect here? Tomkins doesn’t say.
The problem is one of perception associated with the steady diet of evolutionary brainwashing foisted upon us by society. We see a car, computer, or a toaster and immediately comprehend that is has been designed and engineered. However, when we see biological systems that are magnitudes of complexity more highly designed and engineered than the devices produced by mankind, we think that these things arose by chance random processes in some sort of cosmic naturalistic casino. Nothing could be further from the truth – and the data from molecular biology proves it.
Ah, accusations of brainwashing – that seems to be Tomkins’ signature, in the same way that “you can’t have order without God” is Lisle’s.
Remember: design was his premise, and he thinks he has used that to show design. Wonderfully cyclic. Tomkins has a few questions left to answer here, I think.
On a slightly related subject, have you heard about the most recent attack on the Vernanimalcula fossil? If not, you soon will – ENV has already started. Assuming it stacks up we have a case of a supposed fossil of an animal turning out to just be a pattern in the rock. Hilarious.
We really are great at detecting ‘design’ in nature, aren’t we?