More Recent Comments

Friday, October 12, 2007

Eugene Koonin and the Biological Big Bang Model of Major Transitions in Evolution

Eugene Koonin runs a large laboratory at the National Center for Biothechnology Information (NCBI) in Bethesda, MD. (USA) [Evolutionary Genomics ResearchGroup]. His group tends to focus on new ways of analyzing sequences databases and on interesting findings from database mining expeditions.

Koonin and his coworkers are strong supporters of the Three Domain Hypothesis and they usually interpret their data in terms of three domains of life (Bacteria, Archaea, Eukaryotes) with eukaryotes being derived from archaea. As for other evolutonary relationships, Koonin tends to be a "lumper" rather than a "splitter." He will sometimes conclude that two genes or proteins are homologous based on evidence that others find inconclusive.

Creationist Link

Darwin Doubting Heretic Reveals Himself at National Center for Biotechnology
Evolution News & Views
According to Koonin, proteins with similar architecture (folds) are related by common descent even if there's no significant sequence similarity. This is not an uncommon position but it's controversial. Some scientists are not willing to accept that structural similarity alone is sufficient to establish homology. There are too many cases where this assumption leads to awkward and unreasonable implications. Convergence is a possibility that must be entertained.

Koonin has recently published a paper in Biology Direct where he attempts to explain a number of real—or imagined— problems in evolution. In case you're not familiar with this journal, it's a new "open access" journal with an unusual policy. The reviewers must identify themselves and their comments are posted at the end of the paper along with responses from the author(s). Koonin is one of the editors of this journal and he explains the basic philosophy in an editorial [A community experiment with fully open and published peer review].
In Biology Direct, we seek to live by the realities of the 21st century while addressing the issue of information overflow in a constructive fashion and offering a remedy for the ills of anonymous peer review. The journal will publish "essentially anything", even papers that receive three unanimously negative reviews, the only conditions being that three Editorial Board members agree to review (or solicit a review for) the manuscript and that the work qualifies as scientific (not pseudoscientific as is the case for intelligent design or creationism) – and, of course, that the author wants his/her paper published alongside the reviews it receives. Everything in Biology Direct will be completely in the open: the author will invite the referees without any mediation by the Editors or Publisher, and the reviews will be signed and published together with the article. The idea is that any manuscript, even a seriously flawed one, that is interesting enough for three respected scientists to invest their time in reading and reviewing will do more good than harm if published – along with candid reviews written by those scientists. Under the Biology Direct rules, an author is free to solicit as many members of the Editorial Board as s/he has patience for. The philosophy behind this approach is that what really matters is not how many scientists are uninterested in a paper (or even assess it negatively, which could be the underlying reason for declining to review) but that there are some qualified members of the scientific community who do find it worthy of attention. A manuscript will be, effectively, rejected only after the author gives up on finding three reviewers or exhausts the entire Editorial Board. We believe this is fair under the rationale that work that fails, after a reasonable effort from the author, to attract three reviewers is probably of no substantial interest, even if technically solid.
The paper that concerns me here is Koonin (2007) The Biological Big Bang model for the major transitions in evolution. In all cases, the "problems" that Koonin addresses seem to be the rapid and unexplained appearance of novel characteristics, especially those that count as major transitions in evolution. Examples are the origin of cells and the Cambrian explosion.

These events can all be explained by the Biological Big Bang (BBB) model of evolution as Koonin describes in his paper. The concept of the Biological Big Bang is explained in the abstract ...
I propose that most or all major evolutionary transitions that show the "explosive" pattern of emergence of new types of biological entities correspond to a boundary between two qualitatively distinct evolutionary phases. The first, inflationary phase is characterized by extremely rapid evolution driven by various processes of genetic information exchange, such as horizontal gene transfer, recombination, fusion, fission, and spread of mobile elements. These processes give rise to a vast diversity of forms from which the main classes of entities at the new level of complexity emerge independently, through a sampling process. In the second phase, evolution dramatically slows down, the respective process of genetic information exchange tapers off, and multiple lineages of the new type of entities emerge, each of them evolving in a tree-like fashion from that point on.
The BBB model is clearly not gradualistic. Koonin attempts to ally himself with other advocates of episodic change such as Niles Eldredge and Stephen J. Gould. (Where necessary, I have converted Koonin's number references to ones in which the author and date is displayed.)
However, the evolution of life is, obviously, a non-uniform process as described, e.g., in Simpson's classic book [3,4], and captured, more formally, in the punctuated equilibrium concept of Gould and Eldredge [Eldredge and Gould, 1997; Gould and Eldredge, 1993]. Lengthy intervals of gradualist modification are punctuated by brief bursts of innovation that are often called transitions, to emphasize the fact that they culminate in the emergence of new levels of organizational and functional complexity [Maynard Smith and Szathmáry, 1997]
I take issue with Koonin's description of punctuated equilibria (PE). PE is a pattern of evolution that describes speciation events. The morphological changes that characterize a new species are locked in place rapidly during cladogenesis (speciation by splitting). For the most part, these morphological changes are subtle and it often takes an expert to recognize them in the best documented cases. Furthermore, these small changes occur repeatedly in a number of distinct cladogenesis events spread out over tens of millions of years. The result is a very distinctive and well-defined phylogenetic tree that defines the punctuated equilibrium pattern. .

Science Link

Examine macroevolutionary concepts carefully
Nick Matzke
PE has nothing to do with the major transitions that Maynard Smith and Szathmáry discuss in their books (The Major Transitions in Evolution, The Origins of Life). Nor is PE related in any way to the big bang problems that Koonin is addressing in this paper. It's difficult to decide whether Koonin misunderstands punctuated equilibria or whether he is just stretching an analogy. I suspect the former.

It is no coincidence that the Biological Big Bang model borrows terminology from cosmology. Koonin is very explicit about the similarities between his evolutionary model and the cosmological model for the origin of the universe. As a matter of fact, he explicitly addresses this issue in response to one of the reviewers (William Martin) who challenges the comparison. To me, the comparison between biology and cosmology seems forced and I think he weakens his case considerably by comparing the origin of the universe to the Cambrian explosion. Comparisons like that—and the false analogy with punctuated equilibria—contribute to the sense of unease that I had on finishing the paper. One has the distinct impression that Koonin is grasping at straws in order to knock down a strawman.

What are the major problem with evoluton according to Koonin? Are they strawmen? He identifies six major problems and claims that they can all be explained by a model where "evolutionary transitions follow a general principle that is distinct from regular cladogenesis." The BBB is characterized by a phase of rapid evolution with extensive exchange of genetic information between organisms. This phase is followed by a slow phase of evolution of the sort that generates the typical tree pattern.

Before describing the six examples, we need to address the differences, if any, between Koonin's Biological Big Bang and the "net of life" model that is replacing the traditional tree of life at the deepest levels [The Three Domain Hypothesis (Part 5, Part 6)]. The similarities are obvious. In the net of life model the early stages of evolution involved massive exchanges of genetic information such that it is now impossible to construct a traditional tree relating the major groups of species such as prokaryotes and eukaryotes. What Koonin is doing is to generalize this event "by proposing that a phase of rapid, promiscuous evolution might underlie many, if not most of the major transitions in the history of life."

The six transitions are listed below. Each one is followed by a brief comment where I attempt to evaluate its significance.

1. Origin of protein folds
There seem to exist ~1,000 or, by other estimates, a few thousand distinct structural folds the relationships between which (if existent) are unclear.
There is no reason to postulate that all proteins sharing a common fold will share a common ancestor. Some of these proteins might well have arisen entirely independently and evolved to a common fold by convergence. In other words, the underlying assumption that the origin of protein folds represents evolution of some sort may be false. Furthermore, there is even less reason to think that groups with different folds have an evolutionary relationship. In order for this to be true there would have to have been a primordial protein with one kind of fold that gave rise to a protein with another kind of fold. Instead different polypeptides with little three-dimensional structure (random coils) may have independently evolved into proteins with particular folds.

2. Origins of Viruses
For several major classes of viruses, notably, positive-strand RNA viruses and nucleo-cytoplasmic large DNA viruses (NCLDV) of eukaryotes, substantial evidence of monophyletic origin has been obtained. However, there is no evidence of a common ancestry for all viruses.
The reason why there's no evidence of a common ancestor for all viruses may be because there is no common ancestor for all viruses.

3. Origin of Cells
The two principal cell types (the two prokaryotic domains of life), archaea and bacteria, have chemically distinct membranes, largely, non-homologous enzymes of membrane biogenesis, and also, non-homologous core DNA replication enzymes. This severely complicates the reconstruction of a cellular ancestor of archaea and bacteria and suggests alternative solutions.
The existence of distinct bacterial and archaeal domains is hotly disputed. Many groups of bacteria have distinctive features that distinguish them from other groups. There is no need to postulate a radically new mechanism of evolution that accounts for bacteria and archaea if they really aren't much different than the major branches described below.

4. Origin of the major branches (phyla) of bacteria and archaea
Although both bacteria and archaea show a much greater degree of molecular coherence within a domain than is seen between the domains (in particular, the membranes and the replication machineries are homologous throughout each domain), the topology of the deep branches in the archaeal and, especially, bacterial phylogenetic trees remains elusive. The trees conspicuously lack robustness with respect to the gene(s) analyzed and methods employed, and despite the considerable effort to delineate higher taxa of bacteria, a consensus is not even on the horizon. The division of the archaea into two branches, euryarchaeota and crenarchaeota is better established but even this split is not necessarily reproduced in trees, and further divisions in the archaeal domain remain murky.
In addition to eurarchaeote and crenarchaeota there are other groups of prokaryotes that are easily resolved by current techniques (e.g., cyanobacteria, proteobacteria). It may be difficult to resolve the base of the tree because of extensive horizontal gene transfer as postulated in the "net of life" scenarios. This is the one case, along with #3, where the concept of an unusual type of evolution may be correct. I still don't like the term Biological Big Bang to describe it.

5. Origin of the major branches (supergroups) of eukaryotes
Despite many ingenious attempts to decipher the branching order near the root of the phylogenetic tree of eukaryotes, there has been little progress, and an objective depiction of the state of affairs seems to be a "star" phylogeny, with the 5 or 6 supergroups established with reasonable confidence but the relationship between them remaining unresolved.
Substantial progress has been made but the problem is very difficult because of the lack of reliable phylogenetic markers. That, plus the fact that we are trying to sort out events that took place more than one billion years ago. It is too early to conclude that our inability to reach consensus means that something strange must have been going on. That's a cop-out at this time.

6. Origin of the animal phyla
The Cambrian explosion in animal evolution during which all the diverse body plans appear to have emerged almost in a geological instant is a highly publicized enigma [32-35]. Although molecular clock analysis has been invoked to propose that the Cambrian explosion is an artifact of the fossil record whereas the actual divergence occurred much earlier [36,37], the reliability of these estimates appears to be questionable [38]. In an already familiar pattern, the relationship between the animal phyla remains controversial and elusive.
Actually, the relationships between animal phyla are quite well understood at the molecular level. The lineages do not appear to be scrambled by excessive recombination between species as Koonin's hypothesis would require.

The Cambrian explosion is an interesting example of fairly rapid morphological evolution. It may be true that dozens of independent animal lineages simultaneously acquired a new mechanism of evolution during the Cambrian but this does not seem to be the most parsimonious explanation of the data.

To sum up, I don't think that Koonin's examples cry out for explanation in the way he thinks they do. Some of them may not even be examples of evolution. It seems reasonable to attribute the origins of the major groups of bacteria, and the first eukaryotic cells, to a rapid exchange of genes in the beginning phase of life on Earth, but there's no need to postulate that this promiscuous phase was ever repeated at other stages and certainly no reason to assume that there were repeated waves of promiscuity followed by quiescent phases of stabilizaton.

The Biological Big Bang is not so much wrong as it is unnecessary.

Koonin, E. (2007) The Biological Big Bang model for the major transitions in evolution. Biology Direct 2:21doi:10.1186/1745-6150-2-21.

[Photo Credit: The Tree of life is from Ford Doolittle's Scientific American article "Uprooting the Tree of Life" (February 2000). © Scientific American.]


  1. On the other hand, the Origin of Life does cry out for explanation. Have you seen Koonin's take on that?

    Looking back on my post, I see that I was quite gentle. Perhaps I was laughing too hard to say anything mean :P

  2. It may be difficult to resolve the base of the [bacterial] tree because of extensive horizontal gene transfer as postulated in the "net of life" scenarios.

    Of course, it may also be difficult to resolve the interrelationships between bacterial phyla because, as with eukaryotes, the amount of time that has passed since they diverged is enough that further mutations have obscured most of the phylogenetic signal.

  3. That's true but there does seem to be a difference.

    In the case of bacteria, archaebacteria, and eukaryotes, we have seriously conflicting phylogenies depending on which gene you choose. That's consistent with rampant horizontal transfer. We don't see those kind of conflicts with the eukaryotic lineages.

  4. I read this paper yesterday and had similar reactions. The treatment of the Cambrian "explosion" in particular had my eyes rolling around.

    I did find the idea of protein-folding homology interesting (perhaps because, as is probably obvious, I am no biochemist). Seems like it could happen--quasi-neutral amino-acid substitutions could form new hydrogen bonds (or other weak bonds, or even sulfur bridges) that make previous links redundant, freeing up other residues for neutral substitution, etc., eventually obscuring the ancestral sequence homology while retaining the functional shape of the fold. Certainly convergence can't be dismissed either, though, I agree. How could these scenarios be distinguished empirically?

  5. I suspect Larry nailed the problem with Koonin's analogy to inflation. In any case it isn't really appropriate since inflation connotes other physical implications that isn't evidenced.

    What Koonin should reference in his analogy to physics is vacuum selection in general. He does so in the paper on the anthropic principle that Dennehy comments on. I will recycle some of my commentary from PT earlier:

    I may be in error since I have just browsed these papers yet but I think Koonin’s cosmological thinking is confused.

    The weak anthropic principle is used to predict likely values (conditional on the weak anthropic principle), successfully so for the cosmological constant and a few other parameters as of yet. It is not preferred as an open unfalsifiable “just so” description for finetunings or other low likelihood scenarios, as Koonin seems to propose.

    Btw, it is ironical if biologists have started to use vacuum selection scenarios at this time. The earlier mutual support that inflation theory and string theory gave each other for vacuum selection may AFAIU have become problematic with some recent papers. And main stream physics has never taken to it, which of course doesn’t mean it is wrong but that it isn’t a strong contender as of now.

    [Since I wrote this there was a release of a paper that claims bubble nucleation of new universes are so frequent that our own universe are bumped often. We can't notice the imprints that leaves (still unclear exactly what) in later times, but the early ones could be visible and directly test multiverse theories. So vacuum selection may be shored up appreciably again.]

    Seems Koonin has taken to the “a few solid papers, one speculative paper” approach so common among theoretical physicists. That can give some interesting ideas. or laughter. :-P

    But Koonin's large model approach are likely to generate many failures. (Which he confesses to, btw.) And in any case I like my models testable, thank you very much.

  6. Greetings, for the last couple of hours I've been searching about Eugene Koonin and the Biological Big Bang Model of Major Transitions in Evolution and finally I stumble into your blog, it has great info on what I'm looking and is going to be quite useful on my paperwork for the university.
    BTW is amazing how many generic viagra blogs I manage to dodge in order to get the right site and the right
    Thanks for the post and have a nice day