Modern life had a single origin

May 15, 2010 • 10:34 am

If you’ve been reading the evolution websites, you’ll know about the very nice paper in this week’s Nature by Douglas Theobald. (You may remember Theobald as the author of one of the greatest creationism-refuting websites of all: 29+ Evidences for Macroevolution: The Scientific Case for Common Descent. If you haven’t seen it, you should.) In the new paper, Theobald makes a few conservative assumptions to show that the probability that all living species descend from a universal common ancestor is infinitely higher than any other hypothesis, including those of multiple origins of the kingdoms (Bacteria, Eukarya, and Archaea) or of rampant horizontal gene transfer betweeen species that would, by mixing genomes, make life look as though it had a single origin when it didn’t.

Fortunately, I delayed posting on this long enough so that others did the job for me: these include  Nick Matzke at Panda’s Thumb, P. Z. Myers at Pharyngula, and Ker Than at National Geographic.  It’s pretty airtight evidence for evolution, since the hypothesis that trumped all others is that of a single origin of life, with the proteins of existing species showing a pattern of similarity and divergence reflecting the branching bush of evolution.

I’m not sure how creationists will respond to this, but I suppose they could maintain that the data show only that God created life in this way because he needed to give similar proteins to similar species.  That, of course, would require one to believe that those similarities just happen to mimic the similarities expected under evolution.  Proteins group not by lifestyle, but by ancestry. Bats have proteins that resemble those of rats more than those of birds, and whales have mammal-like rather than fish-like proteins.  A marsupial mole has virtually the same niche, and looks almost the same as, a placental mole, but its proteins are more similar to those of a kangaroo.

As P. Z. (and Theobald) point out, the advantage of this new study is that it gives us a number—a probability—with which to gauge the likelihood of modern life descending from multiple origins.

But let us remember (and Theobald mentions this) that we already had incontrovertible evidence for a single origin of all species: the near-universality of the genetic code.

If you’ve studied biology at all, you’ll know that the genetic code—the triplet sequence of DNA (and RNA) that codes for the amino acids of proteins—is virtually identical across all species.  There are 64 triplet codons coding for around 20 amino acids (as well as protein-terminating “stop positions”), and the correspondence between the code and the amino acid is nearly identical across animals, plants, and bacteria.  There are a few exceptions to this, but they are minor: no species deviates from the code by more than a few amino acids, though some mitochondrial DNA deviates by as many as 8 codons.  You can find a list of these deviations (last updated in 2008) at the National Center for Biotechnology Information at the NIH.

Although we can’t attach numbers to the likelihood that the “universal code” reflects a single rather than a multiple origin of life (that would require a model of how the code might have evolved), only a moron or a creationist would deny that this similarity reflects a single origin. There are simply too many genes involved in producing the code and turning it into proteins to think that the code’s universality merely reflects evolutionary convergence in lineages that originated independently. The universality results from ancestry, not coincidence.

In the end, I expect that creationists will find a way around Theobald’s arguments, just as they have around the universal genetic code (Nick Matzke mentions that creationist Paul Nelson has long battled the universal-code argument), but they do so at the expense of their own credibility.

Fig. 1.  The “universal” genetic code (read the triplet code starting with the letter on the left, then the top, and then the right).  Courtesy of Harvard’s Department of Molecular and Cellular Biology.


Theobald, D. L.  2010.  A formal test of the theory of universal common ancestry.  Nature 465:219-223.

24 thoughts on “Modern life had a single origin

    1. We’d need to extend the credibility scale on the negative end. Creationists will, in all likelihood, claim that God made all genes just like we see them, because, hey, noone ever saw an elephant change into a duck.

      If moving the goalposts were part of the ruleset of any known sports, creationists would be the goalkeepers.

  1. I seem to recall a mention somewhere of a study that show some – slight – complexation preference of the codons for the aminoacid they code for. Not much, but presumably enough for selection to work its magic.

    1. Yes, I seem to remember that too, but I believe there where more than a tad “preference” in that test of evolution. 😀 IIRC, similarly to phylogenies you could order the codons with the preference, making it unlikely to be a coincidence.

      And I believe there have been other tests for UGC evolution.

      First, I seem to remember papers where they show that while the code isn’t optimal as regards robustness to reading mistakes such as frame shifts and connected to that cases of (stop) codon/AA/gene elaboration and control, it is very nearly so by a fantastic factor akin to how phylogenies select subsets with such factors as Theobald shows. Presumably the UGC code “froze” while such things were still of paramount importance, for genome size say.

      Second, I seem to remember speculative papers where they connect codons to AA metabolism, and show how it can be elaborated from simpler codings by coevolution. Still, any of those possible pathways tests evolution as an explanation if the results stand.

      Third, I seem to remember likewise speculative papers that ties such coevolution and ultimately metabolism elaboration to natural productivity (occurrence) of AAs. As in, if it is easier to spontaneously produce in nature, it is a likely starting point for metabolism regardless if it was arrived at as a product or adopted as a naturally occurring resource. Again, a test of evolutionary pathway occurrence if the results bear up.

      1. Thanks. I didn’t want to express too much certainty in a vague memory.

        I commented, of course, in the hope that someone would throw out a coupla references.

        1. Yeah, me too. 😀 I know I have them somewhere, but the bookmarks have been reproducing copiously lately, with little selection and much drift.

          However, I found some.

          Here is one on code optimality:

          “Our analysis shows that when the canonical code is tested against a sample of one million random variants using PAM matrix data to measure amino acid dissimilarity, the code appears to be extremely highly optimized at all transition weightings and modular power functions. For the unrestricted set of codes, no better alternatives are found anywhere (data not shown).”

          Here is one on codon ranking vs thermodynamical properties vs natural occurence.

          “Figure 1 – The ranking Robs for the early amino acids (circles) correlates well with the free energy of formation in surface seawater conditions, ΔGsurf. The late amino acids (triangles) have high ΔGsurf. ”

          “In our first approach, amino acids were ranked according to each criterion in Table 1. For example, for criterion M2, glycine (G) is most frequent and is given rank 1. The next two, D and E, are equally frequent, and are both given rank 2.5. The next three are A, V and P, with ranks 4, 5 and 6. All the remaining amino acids are not observed, and are given an equal bottom rank 13.5 (the average of the numbers between 7 and 20).”

          Hope that helps a little.

          1. I should add that the later gives references to the actual codon-AA pairing:

            “A principle component analysis of amino acid physical properties (Urbina et al. 2006) shows that amino acids in the same column cluster in property space,”

  2. Theobald’s enormous probability that life did not have a common ancestor is misquoted (under understated by 180 orders of magnitude) by both National Geographic and PZ. The odds are 10 to the 2,860th power in favor according to the paper (not 10^(2,680) as reported with the 6 and 8 transposed—a mistaken difference of 10^180, itself much larger than the so-called Borel’s Law). Theobald’s paper itself:

    According to a standard objective Bayesian interpretation of the model selection criteria, the scores are the log odds of the hypotheses. Therefore, UCA is at least 10^(2,860) times more probable than the closest competing hypothesis. Notably, UCA is the most accurate and the most parsimonious hypothesis.

    Humorously, Theobald also includes in Table 1 on page 220 the hypothesis that all life except animals and humans have a common ancestor—essentially a test of Genesis 1:20&endash;27. According to the figures reported in Theobald’s paper, the odds of animals arising by separately as in Genesis 1:20 are one in 10 to the 5,264th power, and the odds of humans arising separately as in Genesis 1:27 are one in 10 to the 6,105th power.

    The numbers all come from Table 1, “Class I hypothesis of single versus multiple ancestries” on page 220, in the column ΔAIC, the natural log-likelihood of the probability minus the total number of parameters in the model. Just convert to base 10 these base e likelihoods comparing the common ancestry “ABE” hypothesis to the other noncommon hypotheses:

    “AE+B” hypothesis (most likely uncommon ancestry): 10^(–2,860) = 10^(–6,586/log(10))

    “ABE(–M)+M” hypothesis (common ancestry, except animals): 10^(–5,264) = 10^(–12,120/log(10))

    “ABE(–H)+H” hypothesis (common ancestry, except humans): 10^(–6,105) = 10^(–14,057/log(10))

    I’ll cross-post this comment to PZ as well.

  3. As always when a profound result is teased out of data unexpectedly you amaze at the sheer awesomeness of it all. Here it isn’t profound or startling for evolution, a process which as Theobald shows can admit any natural initial condition, but for resolving deep questions of actual ancestry.

    Thanks! This is also interesting to me since I’m taking a course in Astrobiology with a basis in Earth as a case study. Out of pure interest, in practice verging on hobby I’m afraid, in biology and abiogenesis, if you can believe that. 😀 I’ll be sure to study (paywall!) and mention this when we get to deep ancestry.

    I’ll guess pondering the result without reading the paper et cetera will demonstrate more than my usual naiveté, but that is what I can do at this point. 🙁

    First, it seems to me that this result shows that there is a strength in modeling phylogenies at the molecular level. In fact, AFAIU it shows that LGT, measured at encompassing ~ 1 % of the genome I believe, and endosymbiosis are mere modifiers of such models however large factors they are in reality.

    Then, to reach even further one could perhaps tease out likely pathways for metabolism to shrink the putative LUCA assembly past the RNA genome towards the protobiotic assembly. It should become irrelevant in this aspect out of Theobald’s work, if the modern genome ancestor was merged with metabolism in analogy with endosymbiotic mergers or if it coevolved.

    Second, similarly there is commentary to the fact that this properly embed the UCA/LUCA in an evolutionary population. You can’t properly accept this result while continue to imagine that it all started with a unique cell, as some seem to do at times. (Creationists, I’m looking at you and your Hoyle fallacy!) I like that forcing of the issue as well.

    1. oops! “any natural initial condition” was meant to be “any natural or unnatural initial condition”

  4. … or of rampant horizontal gene transfer betweeen species that would, by mixing genomes, make life look as though it had a single origin when it didn’t.

    But let us remember (and Theobald mentions this) that we already had incontrovertible evidence for a single origin of all species: the near-universality of the genetic code.

    The former argument baffles me, since horizontal gene transfer only makes sense in the context of a near-universal genetic code.

    1. Yeah, nice observation. I believe the opposite hypothesis means that you have to start hand waving about “different modes” at that point. (As in, evolution was somehow under rampant LGT and somehow different.)

  5. What about the possibility that prions are life, and don’t share common ancestry with (D|R)NA life?

    1. Prions are simply proteins, and even those of course “belong” to the species in which they misfire, so they’re not going to show an “independent origin of life.” Moreover, nobody would claim that a prion is alive–at least nobody I know.

      1. This takes us a bit off-topic, since the bit that made me think of it is in a book primarily about life on other worlds, and about strange sorts of life, but here is Peter Ward, pages 21-22 in Life As We Do Not Know It:

        Are prions alive?
        The existence of prions is also a test of the what is a alive question. Prions are small and mysterious “organisms” that cause a number of truly hideous diseases, the most notorious being scrapie (in sheep) and mad cow disease, which can infect humans. When these diseases were first discovered, it was assumed that the infectious agent was a virus. Eventually the “agent” was discovered and named a prion, short for proteinaceous infective particle. Prions test negative for nucleic acids of any kind and thus have neither DNA nor RNA. It is conjectured that they are pure proteins, and there is now much evidence for this protein-only theory of prions. In fact, recombinant prions that are infective to mice have now been made. The mechanism of propagation is believed to involve the ability of the misfolded form of a normal cellular prion protein to cause a shape change in that protein, converting it from the normal shape (harmless) to a new shape that causes great harm to the host, if that is the correct word for the organism containing these “bad” prions. Prions are made/propagated in the cytoplasm in yeast and are thought to be made/propagated on the cell surface in mammals.
        Replicate they do, when they invade nerve cells of their hosts and, like a virus, induce the host cells to start making lots of them. Like the presence of RNA viruses, the existence of prions has interesting implications. They may prove that protein alone can act as both software – coding information – and hardware. Do prions metabolize? Doubtful, since the protein-only theory seems to hold, and it is difficult to envision how a single protein could metabolize. Do they evolve? Very possibly, since the different strains are not equally efficient, and only one will propagate in a single host when a mixture is present, but each will propagate in a single host if introduced serially. Are they alive? Maybe. But like viruses, they cannot be placed into any standard taxonomy of life on Earth.

        Peter Ward isn’t claiming that a prion is alive (and he brings up the main problem with calling prions “alive”), but he is raising the question. I brought it up because it seems to me that the best possibility for life which is not related to all other life on earth would be non-DNA/RNA life. And prions seem to be the closest example now known – at the least, they’re non-DNA, they replicate, and they probably evolve (but probably don’t metabolize).
        (I should perhaps point out that despite what Ward’s last sentence may imply about viruses, in other parts of the book he seems to think viruses are related to other life – after all they use mostly the same codon-to-protein mapping.)

        1. … what is a alive question …

          Oops. Typo not in original (I have a hard bound, so I typed that quote up).

  6. “You may remember Theobald as the author of one of the greatest creationism-refuting websites of all: 29+ Evidences for Macroevolution: The Scientific Case for Common Descent. If you haven’t seen it, you should.)”

    Make sure to also catch Ashby Camp’s rebuttal, Theobald’s counterrebuttal, Camp’s counter-counterrebuttal, and … hey, did Theobald ever reply to this one?

    1. I never heard of that, But it doesn’t matter, you can’t “rebut” actual tests, you have to falsify them. That hasn’t happened.

      1. Or even better, falsify the theory by coming up with a more predictive one. And that is in practice impossible re evolution; too many tests already passed.

        I would say that the likelihood is up there somewhere in “Theobald space”. 😀

      2. May I ask you to be clearer, Mr. Larsson?
        What didn’t you ever hear of?
        And why doesn’t it matter that Theobald’s arguments (as opposed to tests) were rebutted?

        (Note: “rebuttal” need not mean a good rebuttal.)

  7. “only a moron or a creationist would deny that this similarity reflects a single origin. ”

    And here was I thinking that creationists always claimed that the similarity reflects a single origin.

  8. Would this mean that moving, reproducing viruses don’t count as “life”, since they use uracil instead of thymine?

Leave a Reply