Thursday, November 13, 2008

Bad press releases and "revolutions of science

This is a repost of something I just recently posted elsewhere in relation to this crappy article about this press release about this paper. I will stress at the beginning, although I talk about this further down, that I think the paper is pretty good. I don't think it is "earth shattering" or revolutionary, and I detest the way that press releases, even from Universities, tend to use this sort of, frankly inflammatory, language when talking about scientific research. Evolution of evolutionary rates, omptimization processes, etc have all been talked about for awhile in molecular evolution and this paper doesn't completely change how we think about evolution at all. This post also goes into some other things which are tengentially related to talk about this issue before getting into my discussion of the press release and paper itself.

Well hopefully my little diatribe here will help clear some things up. Because it is important to remember that Evolution != Random Mutation + Natural Selection. And it hasn't for quite some time, at least since the synthesis and definitely since the formulation of Neutral Theory. Unfortunately even many non-Creationists and evolution supporters put forward that weakened, watered down, over simplified version as being "evolution" when it isn't. Hell in some surveys even publications like Nature have done it, and it's wrong. People who work in evolution, particularly molecular evolution, have railed against that for a long time. So, with my rant out of the way...

Darwin's original three postulates upon which the theory of evolution was built are simply the following:


* The ability of a population to expand is infinite, while the resources available to sustain said population are finite. This dynamic causes a struggle for existence among individuals as they compete for resources.
* Organisms vary in their physical qualities; these variations allow some members to reproduce more successfully than others.
* These variations are inherited by offspring from their parents.


Note that there is no talk about the source of variation here, and that was one of the major flaws in Darwin's work as perceived by others, he had no mechanistic explanation with how such variation came to be introduced into populations, because he was unaware of Mendel's work. It wasn't until the 1900's, after the re-discovery of Mendel's work that we started talking about genes and mutations, etc. And these three postulates, simplified as they are, still hold true.

For those not aware, there are a lot of different types of mutations including the common point-mutation that everyboyd knows and loves. But there are also larger scale mutations such as insertions and deletions, inversions, chromosomal/genomic amplifications (including the subset of gene duplications), gene fusion events, chromosomal rearrangements, etc. Setting that groundwork is very important because the diversity of what we are talking about in terms of mutations is pretty broad and capable of some pretty amazing things. Gene duplication followed by divergence and neo-functionalization is something very important in terms of the evolution of new protein functions and surprise surprise adds new information to the genome. Gene fusion events happen and so you can end up with two formerly interacting genes now being one gene, inversions can re-arrange the order of protein domains while leaving the protein functional, etc. There is a lot going on.

The Neutral Theory of Molecular Evolution was introduced by Kimura in the late 1960's. Unfortunately early on some people took it as being an argument against selection and evolution although it never was, Kimura was carefult o stress that it was complimentary. There are still debates about the relative importance of selection versus drift but neutral evolution today is well accepted, particularly among molecular evolutionary biologists, and neutrality typically is our null hypothesis in terms of selection. Neutral theory illustrated the fact that mutations are not just beneficial or deleterious, there are also two other classes of mutations: neutral (no selective effect) and nearly neutral (weak selective differences either positive or negative). Selection and Genetic Drift are oposing forces in evolution. Drift is a purely random process introduced by stochastic sampling biases in sexually reproducing organisms. It can be quite powerful in terms of resulting in the fixation or loss of allelic variants in a population in the presence of even weak selection but most particularly when selection is non-existent. This can be a very powerful force in terms of driving speciation. In fact in Peripatric Speciation drift is thought to be a very important factor due to founder effects and population bottlenecks. For anyone interested in genetic drift, speciation, mutation, and how allele frequencies change in populations you may want to look into the field of Population Genetics, a sub-discipline (essentially) within evolutionary biology that studies population level effects on allelic frequencies.

Now, back to the paper at hand. The Princeton press release is better than the linked article by far, although I still take issue, as usual, with lots of the language used in these sorts of press releases. "Until now evidence of this was lacking", etc. They all irritate me. I think the work done by this lab was important, but it isn;t revolutionary. It is another really well laid out example of some of sorts of things that have been talked about and studied in molecular evolution for awhile. here is the link to the actual research article which you should be able to view if you have academic access to the journal. This paper is interesting and is good experimental evidence, and a reuslting analytical framework for what they call control optimization, in this case of redox potentials in the electron transport chain (Cytochrome b in particular). It isn't, in my opinion, revolutionary. Press releases are frequently really, really bad and tend to overhype findings and put them in the context of "changing everything we know" which simply isn't true. That said at a glance this seems like pretty good and interesting work and I'm going to read it in much more detail when I can, in particular because there may be something there that helps me in my own work, maybe. We'll see. So I just want to talk about a few other things in this post as well.

Compensatory mutations have been known for a long time. If you have a coding sequence for a protein and let's say a mutation in site i of that protein which results in the change from a relatively large amino acid, such as tryptophan, to a small one like alanine. Now site i was in interaction with site j, which had a relatively small amino acid at that position. Due to the mutation the interaction between these two amino acids is weaker or perhaps non-existent which, for the sake of argument we will assume has an effect on the function which makes the protein less efficient but doesn't result in the death of the organism or anything very drastic. Now, over evolutionary time any mutation nat site j which restored or strengthened that interaction will be beneficial because it improves function, this may even happen relatively quickly because the selective advantage of the compensatory mutation is quite high and so it may become fixed in the population quite quickly. Those are compensatory mutations and we've known and talked about and studied them for quite some time in molecular evolution.

Things like co-evolution and co-evolutionary rates have also been known for awhile, with many studies, such as this one being some examples of that sort of thing. The take home message is that functionally related proteins and, more importantly, functionally interacting proteins, tend to have correlated evolutionary rates. It is a simple extension of both neutral theory and natural selection which is actually somewhat intuitively obvious.

For those interested I thought I would also provide a link about the evolution from RNA to DNA. Why? Well because it goes along with my next topic which is the molecular repair machinery. So DNA repair effects things like mutation rates, because it fixes lots of mutations before they can be passed on to the next generation. But as you are reading about them keep in mind a few things.

1) Repair mechanicms are not perfect, like all molecular mechanisms they are "leaky"

2) Depending on the type of damage being repaired they can actually introduce changes themselves. For instance if there is a base mismatch between the two strands of DNA the repair mechanism has only a 50/50 chance of repairing it correctly, it doesn;t KNOW what the coding strand is, it just fixes the mismatch pretty much at random. So it may repair the mismatch correctly or, in the repair process, actually introduce a mutation in the coding sequence.

3) The DNA repair machinery is coded for on the DNA itself, meaning that it is also subject to mutation and selection. That means that the rate and accuracy of DNA repair mechanisms play a role in determining the rate of evolution itself. Leaky or error-prone repair mechanisms tend to result in a higher rate of mutation and thise a higher rate of evolution, stricter and more accurate repair mechanisms result in the opposite. It's a fine line and trade-off and there is ample evidence of selection on DNA repair mechanisms, particularly among mammals. A good review of the evolution of DNA repair mechanisms themselves can be found Here, although it is a journal article and you may need an academic subscription to the journal (if you are a university student or staff you should be able to access it automatically if on a university network computer or you can access it through your libraries electronic resources.

Papers like this one also talk about this sort of thing.

Hope that helps and wasn't too long or dull!