Casey Luskin, a blogger for the Discovery Institute, recently took issue with my proposal that when reading a scientific paper, one should check the institutional affiliation and credentials of its authors. I believe the part that he particularly objected to was my statement that one might not wish to use the Discovery Institute as a “scientific authority on evolutionary theory.”
“In other words, study a paper carefully, but if the authors work with Discovery Institute, disregard everything they are saying from the outset. That’s the ground rule that comes before any other tips. It’s a great way to keep yourself carefully in the dark about things you know nothing about. And she calls us “agenda-driven”?
Imagine how journal editors would behave if they followed Raff’s advice. Or better yet, imagine what would happen if Raff herself were a journal editor. Someone affiliated with Discovery Institute (or any group friendly to ID) submits a paper, and you immediately toss it in the trash without even taking it seriously. More than a few such editors probably share her philosophy. That doesn’t exactly inspire confidence in the peer-review system, even though of course there are already plenty of reasons to lack such confidence.”
It’s a very telling reaction on his part. The Discovery Institute insists that Intelligent Design research is scientifically undermining the theory of evolution, and they are obviously very touchy about perceived criticism of their scholarship. I would remind Mr. Luskin that that the US Federal court in Dover PA has ruled on presented evidence that Intelligent Design is not science and has not contributed any new research on the subject of evolution:
We find that ID fails on three different levels, any one of which is sufficient to preclude a determination that ID is science. They are: (1) ID violates the centuries-old ground rules of science by invoking and permitting supernatural causation; (2) the argument of irreducible complexity, central to ID, employs the same flawed and illogical contrived dualism that doomed creation science in the 1980’s; and (3) ID’s negative attacks on evolution have been refuted by the scientific community. As we will discuss in more detail below, it is additionally important to note that ID has failed to gain acceptance in the scientific community, it has not generated peer-reviewed publications, nor has it been the subject of testing and research. Expert testimony reveals that since the scientific revolution of the 16th and 17th centuries, science has been limited to the search for natural causes to explain natural phenomena. (9:19-22 (Haught); 5:25-29 (Pennock); 1:62 (Miller)).
It seemed to me that the best response to Mr. Luskin (and various creationists calling me a liar in the comments on my blog) is to do exactly what they’re afraid I won’t: take an Intelligent Design research paper seriously and subject it to the same rigorous criticism that I use in any other paper that I review for journals. Let’s see if it holds up to scrutiny.
The Intelligent Design community’s current flagship journal is called “BIO-Complexity”. It aims “to be the leading forum for testing the scientific merit of the claim that intelligent design (ID) is a credible explanation for life.” BIO-Complexity “publishes studies in all areas of science with clear relevance to its aim, including work focusing on the relative merit of any of the principal alternatives to ID (neo-Darwinism, self-organization, evolutionary developmental biology, etc.).”
I chose to read the most recently published (2012) research article (“A Tetrahedral Representation of the Genetic Code Emphasizing Aspects of Symmetry”). The author was a postdoc at Baylor College of Medicine (a legitimate research institution) at the time of the paper’s publication, although he doesn’t seem to be affiliated with them any longer.
(Incidentally, this is a perfect opportunity for people to exercise their critical reading skills. I invite anyone who’s interested to look through my guide to reading scientific papers, and actually try it for yourself on this example paper before you go any further in this post. Then, compare your interpretation with mine, and let me know what you think in the comments below!).
In order to understand this paper, you need to understand how proteins are made. Don’t let this daunt you! Remember, Mr. Luskin isn’t a scientist either, he’s a lawyer. And I have to assume, given his position, he’s read this article and understood it. So you can too, with a little bit of background. Here is the process broken down very simply:
1. Different kinds of proteins have different shapes.
2. Each protein is made up of chains of different chemicals, called amino acids. There are 20 amino acids, and they can be strung together in any combination the cell dictates.
3. The order of amino acids present in the protein determines its structure. Each amino acid has a different structure and chemical properties, so different combinations of amino acids lead to different types of proteins.
Simple, right? Now, how does the cell specify the order of amino acids?
4. The order of amino acids is determined by a coded message. This code is carried by a strand of RNA, called messenger RNA (mRNA). mRNA signals to the cellular machinery which amino acids should be placed in a particular order, forming a long chain.
[Bonus information: There are actually two levels of coding! mRNA itself is coded for by DNA. That’s what we mean when we refer to your “genetic code”. When particular genes are “switched on”, they direct the cell to make mRNA, and then proteins. When they’re “switched off”, they don’t make mRNA, and therefore no protein. ]
Here is short video showing an overview of the process:
If you want to really understand this concept well, this (longer) video aimed at teenagers is an incredibly good demonstration/explanation of how the DNA code is used to make protein (attention, parents and teachers! This is a good teaching tool):
And here is the best source you could possibly read on the subject (an excerpt from Stryer’s Biochemistry), though it’s very technical. (hint: start on page 17).
So now that you know the general process of how proteins are made, let’s focus in on one small step: How does mRNA code for amino acids?
If you think back to basic high school biology class, you’ll remember that DNA has four bases: A, C, G, T. Their order makes each gene different. RNA also has four bases: A, C, G, but instead of a T it has a U. So an mRNA strand might look like this:
It turns out that every three bases code for one amino acid, and every combination has been worked out. So we know that ACU calls for the amino acid threonine (thr), and GGA calls for glycine (gly), for example. We even have a table that shows it. Using the table, can you figure out the amino acid chain from my example above?
There’s something special about the first and last triplets in my example above. They’re actually important signals (“start”/Met and “stop”) to the protein-making machinery (called “ribosomes”) that define the beginning and end of the amino acid chain.
Now if you’re paying attention, you’ll notice that there are only 20 amino acids, but actually 64 possible three-base combinations (we call those “codons”). So there are some amino acids that are specified for by multiple codons. Take another look at the table to see what I mean. There’s a special biological term for that: degeneracy. Degeneracy doesn’t mean the same thing as it does in our day-to-day usage. Instead, it simply means that some amino acids are coded for by more than one codon (synonyms). Interestingly, it tends to be only the last base that varies between synonyms. (There are important biochemical reasons for this that I won’t get into, but you can find more about it in the Stryer chapter I linked to above).
Got all of that? Okay, let’s move on to the paper!
As I always do, I skipped the abstract and began with the introduction, to avoid biasing my interpretation. First, I identified the big question of the article:
BIG QUESTION: Remember, this is not the question the paper is trying to solve, but the question that the field is trying to solve. In this case, my identification of the BIG QUESTION would therefore be: What is the scientific evidence for an Intelligent Designer?
Next, I looked at the background provided by the author. While I’ve already summarized most of it above, I want to call your attention to something that I would criticize any author of any paper I was reviewing: there are no citations to any previously published work for the statements that he makes in his introduction. The only citations he presents are those to other proposed amino acid tables. It may seem like a silly thing, but it’s actually very important. All work that’s not original to a paper must be cited. Failure to do so is a red flag on many levels; the author, the reviewer, and the journal editor all should have caught this.
The author goes on to discuss amino acid tables, such as the one I posted above and notes that they’re “convenient representations of the raw facts of correspondence, but they also point to the underlying order of the code.” (emphasis mine).
As I read this sentence, I immediately stopped and asked: What does the author mean by “underlying order”? Order in this context implies intentionality and design beyond “raw facts of correspondence” (these codons specify those amino acids). Has this been demonstrated experimentally in some way? If so, who did the work, and how did they do it? It’s an important premise that the author builds the foundation of the paper upon, but it isn’t justified by any citation to previous work. He just throws it out there, using as his only example the fact that codons with uracil in the middle position encode hydrophobic amino acids.
If the author is using this as evidence of “underlying order”, he needs to clearly explain how this demonstrates it. It’s true that the sixteen codons with uracil in the middle position encode hydrophobic amino acids, but if this is due to some kind of “underlying order”, why aren’t the equally hydrophobic amino acids tryptophan or cysteine also encoded by a middle-position uracil? Without giving any additional explanation or citation to experimental work, it appears that he’s hunting spurious patterns with no biochemical justification.
The author then offers the following:
“If this is an organizing principle, might there be others as well? That possibility, coupled with the idea that some geometric representations of the code might display the code’s underlying organization better than others has generated interest in functional representations—ways of displaying the basic mapping that emphasize possible principles of organization rather than ease of looking up the mapped pairs. If the genetic code can be represented in ways that offer important insights into the biological properties of its components, these representations may be of use both in education and in bioinformatics.”
So, in other words:
1. The uracil-hydrophobic pattern (again…what pattern???) shows that there’s an “organizing principle” to the amino acid code.
2. If there are other principles, maybe rearranging the genetic code in some specific way will help researchers find them.
From this, I would identify the author’s specific question (what the paper is about) as:
What are organizing principles of codon usage?
I’d be interested in seeing some kind of analysis that might establish why uracils in the center position encode some hydrophobic amino acids, but not others, for example. The author doesn’t cite any papers on this subject, so I would expect that he might propose some hypotheses for this and then do the experiments to test them.
EDITED TO ADD: We actually do have some evolutionary explanations for why the amino acid code is the way it is. See for example the chapter from Stryer that I linked to above, or this reference, or this one. (H/T to rbeagrie and Matt Hodgkinson for sending me links to those in the comments).
But neither these, nor any other papers on that subject, were cited.
Instead he proposes the following approach:
(This research) will try to find a different, 3 dimensional geometric representation of the code that might better display its organization with an emphasis on function of amino acids.
In his own words:
“The idea motivating this work is that a geometric representation of the code will only be as compelling as the harmony between the chosen geometry and the biological reality.”
The author just wants to reorganize the standard coding chart? How does that answer the specific question? We already know which codons make which amino acids, and their chemical properties. Does reorganizing the standard chart actually provide us with any new information about the “organizational principles” of codon usage?
Moving on to the methods, with these questions in mind…
The author has decided at the outset that a tetrahedral representation of the amino acid code is the best choice, since:
“A tetrahedral representation of the genetic code should fit naturally within this geometry if it is to illustrate real underlying order. Otherwise it would merely be an instance of displaying the code on a shape that does not fit it in any compelling sense”
The author explained his choices for how to construct the tetrahedral representation: repeated rows of equilateral triangles, with each cell of triangle representing a codon. Start and stop codons were placed at the vertices of the triangle. He chose to fill in the remaining cells with amino acids according to the following principles: 1) place the hydrophobic amino acids in the central cell (presumably because that’s where hydrophobic amino acids go in proteins), 2) place the others in cells according to principles of “balance and symmetry” (a rather subjective approach) and/or function.
And that’s about it. There are some diagrams to illustrate the different ways in which the author rearranged the amino acid table, and compared to previous representations. For example, here is Figure 7:
(Compare this with the standard table I’ve posted above. Do you find it easier to use? Or can you derive any new insights about the underlying organizational principles from it?)
So, to summarize the methods, the author simply rearranged the table of which mRNA codons specify which amino acids, grouping amino acids by function. There’s not much to it, other than to utilize “principles of balance and symmetry” to create an aesthetically pleasing pattern.
The author concludes that he has “attempted to construct a geometric representation of the genetic code that emphasizes the natural patterns of symmetry and periodicity.” The author states that his method of tetrahedral representation is superior to another researcher named Fujimoto [Note: Fujimoto’s paper should have been presented as background information, NOT brought up first in the discussion] because Fujimoto’s “does not aim to represent the kind of underlying order that has been the focus here, namely the natural patterns of symmetry and periodicity.” [At this point I got curious what Fujimoto’s representation actually was, and what he said about it, so I looked up the reference. Turns out it’s to a patent, not a scientific article. Indeed, Fujimoto doesn’t say anything at all about the codon usage representing natural symmetry. The patent is merely presented as a device to aid in looking up/memorizing the genetic code.]
The author reiterates how he used the principles of balance and symmetry, and grouped amino acids by properties, in order to devise his representation. He then discusses the significance of his work. I’m just going to let his own words speak for themselves:
“The final tetrahedral representation presented here is therefore offered not as a demonstration of any new facts, but rather as an application of existing facts, the potential significance being that this way of organizing them may provide new insights.”
In my judgment, this is not a rigorous paper, and would almost certainly not be published by one of the more widely read biological journals. The only citations in the background material are to previous representations of the genetic code. There’s a fundamental assumption (that the codon code reveals principles of “underlying order”) built in to the paper, which is presented as a simple assertion, with no reference to any research justifying it. The methods are highly subjective. On final analysis, this paper isn’t really anything more than a rearrangement of existing information, trying (and failing) to find meaningful patterns in it. No new scientific knowledge has been the result. No further hypotheses were proposed. No discussion was provided of what future work might be done. The author himself even points this out.
[For a contrasting example of research involving the amino acid code, here’s a forthcoming paper (and news summary)that looks at how less frequently used codons might affect protein production. In this project, the authors actually tested their hypotheses—they created more than 14,000 mRNAs and analyzed the effects of different codon usage on protein production. Hopefully you can see the difference between the approaches of these two papers.]
So, Mr. Luskin, rather than “tossing it in the trash” unread, I have given serious consideration to the most recent research paper that I can find from a journal affiliated with the ID community. I’m not impressed. I stand by my position that the Discovery Institute wouldn’t be a good choice when seeking expertise in evolutionary biology. My cautions with regard to Discovery Institute and ID papers isn’t because I can’t stand anyone who disagrees with me. Rather, the Discovery Institute and its affiliated authors have again and again shown themselves incapable of producing any meaningful, rigorous, and original research that challenges evolution. The responsibility to establish scientific credibility rests with your side.
Many thanks to Rudy Raff and Steve Scott for comments on drafts of this.