The Tree of Life – Part 2 of 2

The world’s foremost atheist (not much of an accomplishment) and most famous writer on evolution is Richard Dawkins. In a recent interview, he was asked,

Out of all the evidence used to support the theory of evolution, what would you say is the strongest, most irrefutable single piece of evidence in support of the theory?[1]

Dawkins replied that it is difficult to pick out the best evidence, since there is so much of it, but he eventually settled on genetic evidence, the sort that we examined in part 1 of this post, in which proteins or genes of different organisms are compared:

There’s an enormous amount of evidence, from all sorts of places, and it’s hard to pick one strand which is more important than any other… I think to me perhaps the most compelling evidence is comparative evidence, from modern animals – particularly biochemical comparative evidence, genetic, molecular evidence.

After briefly explaining the methodology, Dawkins presents his conclusion:

So, you can take any pair of animals you like – kangaroo and lion, horse and cat, human and rat – any pair of animals you like, and count the number of differences in the letters of a particular gene, and you plot it out, and you find that it forms a perfect branching hierarchy. It’s a tree, and what else could that tree be, but a family tree.[2]

Well… baloney. As we’ve seen, evolutionary trees constructed on the basis of genes contradict each other all the time. So it’s time to let Dr. Jekyll speak again. Churakov et al., discussing precisely the example mentioned by Dawkins, found that five genes supported a pattern in which human and armadillo are most closely related; nine genes supported a contradictory history, where human and elephant are most closely related, and eight genes supported a third (hi)story, where elephant and armadillo were closest relatives.[3] And writing in the journal Evolution and Development [4], researchers entitled their paper Conflicting phylogenetic signals at the base of the metazoan tree, and wrote in the summary that “… analyses of… sequences fail to resolve the relationships… We demonstrate that… conflicts in the phylogenetic signal contained in different amino acid sequences obscure the phylogenetic associations among the early branching Metazoa. These factors raise concerns about the ability to resolve the phylogenetic history of animals with molecular sequences.”

Genes vs. Bones

The journal Nature weighed in with a review article that spoke about evolution wars.[5] This was not a reference to conflict between biologists and critics of evolution. It was a reference to wars within biology. One problem, as we have seen, is that there is a huge discrepancy between different trees-of-life, depending on which protein or gene you use. But there is also a divergence between tree-of-life models when you contrast those made from anatomical evidence and those made on the basis of genetic evidence. It’s as if you compare two computers. If you look at the hard-drive, the evidence indicates that one computer came from South Korea and the other from Taiwan. But if you look at the motherboard, you conclude that one computer came from Oshkosh and the other from Tierra del Fuego. The widespread disagreement between molecule-based evolutionary trees and anatomy-based evolutionary trees led the Nature writer to comment that “Evolutionary trees constructed by studying biological molecules often don’t resemble those drawn up from morphology [the study of the structure and anatomy of the organism].”

Despite these clear statements in the professional literature, evolutionary biologists persist in overselling their brand. Notwithstanding the research that contradicts their absolutist claims, they often insist, when writing for the broad public, that molecular studies confirm anatomical studies. For example, in his book Galileo’s Finger, Oxford University scientist Peter Atkins boldly states, “The effective prediction is that the details of molecular evolution must be consistent with those of macroscopic evolution.” He then adds, “That is found to be the case: there is not a single instance of the molecular traces of change being inconsistent with our observations of whole organisms.”[6] Yet a variety of reports clearly recognize that these studies frequently conflict with one another. One authoritative review paper by Darwinian leaders in this field stated, “As morphologists with high hopes of molecular systematics, we end this survey with our hopes dampened. Congruence between molecular phylogenies is as elusive as it is in morphology and as it is between molecules and morphology.”[7] Another set of pro-evolution experts wrote, “That molecular evidence typically squares with morphological patterns is a view held by many biologists, but interestingly, by relatively few systematists. Most of the latter know that the two lines of evidence may often be incongruent.”[8] And despite consistent attempts by apologists to portray developments in this area of biology as something less than a crisis for evolutionary theory, the news is finally making its way into the popular press. In 2009, for example, The Telegraph reported that, “Charles Darwin’s tree of life is ‘wrong and misleading’”.[9]

In July 2010, a Johannesburg educator contacted a distinguished American geneticist and posed a number of questions about evolution to him.[10] The educator has no training in science and little knowledge of biological evolution. He was caught in the cross-fire of the controversy regarding the teaching of evolution in Torah schools and was seeking some enlightenment. In his response, the American biologist played the perfect Mr. Hyde:

There is overwhelming evidence for evolution. The more we learn and the more powerful our technologies become, the greater our insight is into the relationship of all living organisms, past and present.

This statement came with no qualification: nothing about contradictions, difficulties, wars, absurd classifications, incongruities, “burial” or “annihilation” of the tree of life. Just that we have overwhelming evidence for evolution on the basis of advanced research into relationships between organisms. To say that the professor was selling lokshen to the teacher is to insult pasta.

Paradigms

The typical reader, confronted with this material, might think that with so many setbacks, evolutionary biologists would at least countenance a different view of reality. The typical reader, alas, is not familiar with the mechanics of scientific paradigms. Thomas Kuhn was arguably the most influential philosopher of science in the twentieth century. In The Structure of Scientific Revolutions, first published a half-century ago, Kuhn did a marvellous job in describing how the most productive work done by the vast majority of scientists is in solving puzzles within the regnant ideology.[11] Little attention is paid to all the loose ends – those parts of the paradigm which are not explained and, indeed, unexplainable. It is only when the difficulties accumulate to the point where the theoretical structure of the theory collapses that the bulk of the community of scientists acknowledges the problems.

In other words, paradigms are much like flypaper: once stuck, it’s hard to leave. The flood of counter-evidence has not caused biologists to desert the evolutionary paradigm en masse. The lingo changed – where before biologists spoke of the Tree of Life, they now often speak of the web of life or the bush of life [12] or the mosaic of life. Or rather, when they speak to each other they use the new terminology. In pronouncements to the public or in textbooks, it’s the same old story: the tree-of-life is this impossibly gorgeous, perfectly consistent research programme that confirms the Darwinian worldview without a wrinkle.

Paradigms often take at least a generation to overturn. In a trenchant insight into the workings of modern science, the great physicist Max Planck, the father of quantum mechanics, remarked:

A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.[13]

Still, there is hope. In the New Scientist article cited above, evolutionary biologist Eric Bapteste writes, “If you don’t have a tree of life, what does it mean for evolutionary biology? At first it’s very scary – but in the past couple of years people have begun to free their minds.” And philosopher of biology John Dupré adds that it is all “part of a revolutionary change in biology. Our standard model of evolution is under enormous pressure.” The belief that all species are related and evolved from common ancestors is sacrosanct among many biologists. No doubt, many will stick to the dogma like a tattoo to a biker. But the Tree of Life has fallen with such a thunderous thud that it is forcing some biologists to consider different possibilities. A paper published in August 2011 by four European evolutionary biologists argues that the Tree of Life is “becoming increasingly implausible.” Although the Tree of Life “has been stretched to fit the data” in various ways, “given our knowledge of the data, it seems that the elastic limit of the original hypothesis has been passed.”[14] Some even go so far as to acknowledge that the whole field has made little progress in the past century: a review article in the journal BioEssays reported that despite a vast increase in the amount of data since Darwin’s time, “our ability to reconstruct accurately the tree of life may not have improved significantly over the last 100 years.”[15]

It is quite entertaining to watch researchers swinging from tree to tree, first concluding that they have found The One True Tree, and then, after obtaining contrary results, ditching the first tree and replacing it with a quite different specimen. We could continue to discuss the crises befalling the Tree of Life paradigm, but perhaps it’s best to end by noting that in Stevenson’s novel, the affable Dr. Jekyll is the awful Mr. Hyde. He is so tormented by his inability to stop metamorphosing into his monstrous alter ego that, in desperation, he finally commits suicide.

Concluding Remarks

Those who jump into the evolution debate often drown under the claim that there is overwhelming evidence for evolution. It is a claim that is meant to preclude any possibility of challenging the validity of evolutionary biology. Any instance of counter-evidence is brushed aside, often without careful consideration, because, “Hey, there is overwhelming evidence for evolution.”

My experience has been that those who claim that there is overwhelming evidence for evolution are like the schoolyard bully who always threatens to open up with a machine gun but, when push comes to shove, produces a pea-shooter. The invincible evidence for evolution – when actually chewed rather than swallowed – is far more vulnerable than is made out to be.

Richard Dawkins, widely considered the most influential spokesman for evolutionary biology, stated that genetic analysis constitutes the strongest, most irrefutable evidence for evolutionary biology. The expectation, going back several decades, was that common descent would be smoothly vindicated by genetic data. It would corroborate (or clarify, as the case may be) evidence from anatomy and palaeontology regarding the relationships between all organisms, and would once and for all convince everyone that we are all descendents of bacteria.

This expectation has turned out to be wrong. In a classic case of piling one epicycle atop another, researchers committed to the evolutionary paradigm twist themselves into loops trying to morph trees into bushes, and bushes into webs, and webs into mosaics or whatever the latest metaphor happens to be. The truth is that genetic analysis is a Rorschach test. Evolutionary biologists see relationships because they want to see relationships. The exercise starts within the Darwinian paradigm, and pre-supposes that all organisms are biologically related. But when you allow yourself to look past the evolutionary horizon, you see that the genetic data does not support a picture of common descent. Evolutionary biology is a sartorially-challenged emperor. It’s time to depose him.

References:

[1] http://www.youtube.com/watch?v=5PlqNoCAIgA.

Retrieved 13th February 2011.

[2] See http://www.evolutionnews.org/2011/02/for_darwin_day_false_facts_and043691.html.

Retrieved 12th February 2011.

[3] Mosaic retroposon insertion patterns in placental mammals, Gennady Churakov, Jan Ole Kriegs, Robert Baertsch, Anja Zemann, Jürgen Brosius, and Jürgen Schmitz. Genome Research, Cold Spring Harbor Laboratory Press, March 2009. The article can be viewed here: http://genome.cshlp.org/content/19/5/868.abstract.
Retrieved 13th February 2011.

[4] EVOLUTION & DEVELOPMENT 5:4, 346–359 (2003). See the article here: http://people.vanderbilt.edu/~antonis.rokas/pdfs/2003_Rokas_etal_Choanos_EvolDev.pdf.
Retrieved 13th February 2011.

[5] Trisha Gura, “Bones, Molecules or Both?,” Nature, Vol. 406:230-233 (July 20, 2000).

[6] Peter Atkins, Galileo’s Finger: The Ten Great Ideas of Science, page 16 (Oxford University Press, 2003).

[7] Patterson et al., “Congruence between Molecular and Morphological Phylogenies,” Annual Review of Ecology and Systematics, Vol 24, page 179 (1993).

[8] Masami Hasegawa, Jun Adachi, Michel C. Milinkovitch, “Novel Phylogeny of Whales Supported by Total Molecular Evidence,” Journal of Molecular Evolution, Vol. 44, pages S117-S120 (Supplement 1, 1997).

[9] http://www.telegraph.co.uk/science/4312355/Charles-Darwins-tree-of-life-is-wrong-and-misleading-claim-scientists.html.
Retrieved 15th July 2012.

[10] See https://torahexplorer.com/2010/07/04/readers-feedback-professor-james-shapiro-rabbi-blue/.

Retrieved 6th August 2012.

[11] Thomas Kuhn, The Structure of Scientific Revolutions, third edition, The University of Chicago Press, 1996.

[12] Here is one example of the genre. An article in Trends in Ecology and Evolution concluded, “the wealth of competing morphological, as well as molecular proposals [of] the prevailing phylogenies of the mammalian orders would reduce [the mammalian tree] to an unresolved bush…” W. W. De Jong, “Molecules remodel the mammalian tree,” Trends in Ecology and Evolution, Vol 13(7), pages 270-274 (July 7, 1998).

[13] Planck (1949) Scientific Autobiography, pages 33-34.

[14] See http://www.biology-direct.com/content/6/1/41/abstract.
Retrieved 5th September 2011.

[15] Matthew A. Wills, “The tree of life and the rock of ages: are we getting better at estimating phylogeny,” BioEssays, Vol. 24: 203-207 (2002), reporting on the findings of Michael J. Benton, “Finding the tree of life: matching phylogenetic trees to the fossil record through the 20th century,” Proceedings of the Royal Society of London B, Vol. 268: 2123-2130 (2001).

Advertisements

6 Responses to “The Tree of Life – Part 2 of 2”

  1. this is a test message Says:

    I like the beneficial information you provide you with with your content.I will bookmark your web site and look at once again the following on a regular basis.I am rather certain I will learn about loads of new stuff best suited the following! Great luck to the future!

  2. Matt Says:

    One of the major arguments made in this two part series is as follows:

    Yoram makes a claim that if evolution by descent with modification were true, then it must be the case that *every single* individual gene can be used to create a phylogenic hierarchy (“Protein or gene comparison is valid, if at all, only if it is consistent regardless of which genes or proteins were used in the comparison”). If that is not the case, he argues, it would “disprove evolution” in some sort of Sherlock Holmes “aha!!” sort of way (that rarely happens in science). He then goes on to present what is called the single-gene discordance problem: that using single genes can produce nonsensical and contradictory results *sometimes*. Thus, he argues, the whole idea of common ancestry falls into question.

    There are several problems with this line of argumentation, which I will deal with in the following comments.

  3. Matt Says:

    ..Which brings me to PROBLEM 2:

    Yoram fails to mention that the literature offers many solutions to the problem of single gene discordance problem, and by authors prominently feature in his own bibliography:
    Rokas et al (Nature 425, pp798-804, 2003) show that, while single genes yield discordant results, suites of small numbers of randomly chosen genes do in fact produce consistent species trees “with remarkable support.” If Yoram wants to really question genetic relatedness, no number of genes should produce consistent phylogenies. “In this phylogenetic analysis of eight yeast taxa, concatenated data sets of 20 genes are sufficient to provide very strong (>95%) support for the species tree” (ibid). “The results demonstrate that concatenation of a sufficient number of randomly selected genes overwhelms conflicting signals present in different genes” (ibid). Rokas is not the only researcher to demonstrate this. And yeast taxa are by far not the only branch of the tree of life to demonstrate consistent, repeatable genetic trees. In my brief survey of the literature, I have found that the bulk of papers provide simple, well-motivated algorithmic modifications that produce taxonomies that are (a) consistent and (b) agree with other phylogenic approaches.

    In short: Of course single genes do not produce clean taxonomies, but that isn’t the end of the story. Focusing on one problem, while ignoring the explanations, and solutions is not proper scientific skepticism. Focusing *only* on the anomalies, while ignoring the majority of successful phylogenetic reconstructions is likewise improper.

  4. Matt Says:

    To summarize,

    Yes, single gene methods do not produce consistent trees, but

    1) These are to be expected.
    2) Many of the reasons for this discordance are well understood. This is a work in progress.
    3) Many working solutions have been offered.

    Single-gene methods are not the only genetic methods for constructing taxonomies. Many other other genetic methods have proven very successful in producing well-understood taxonomic relationships and produced some unambiguous demonstrations of inheritance from common ancestry.

    Science is not perfect and fraught with places where tidy theories intersect with the messiness of the real world. What a proper informed scientific consumer seeks to do is ask: do multiple lines of evidence using complementary methods robustly demonstrate the principle in question, despite that messiness? To the extent that there are methodological problems and confounding affects (and there will be), can those effects be understood and addressed. Is the science moving towards a better understanding of these challenges or just moving around in circles? Those whose sole interest is in creating doubt and muddying up the water will just pluck the experimental challenges and uncertainties of science out-of-context and string them together in broad rhetorical arguments. They cherry pick quotes and sources only to the extent that those quotes reinforce what they wanted to hear. They look no further and no more rigorously beyond that.

  5. Yoram Bogacz Says:

    I will respond to Matt’s comments in a post entitled More on the Tree of Life, G-d willing. Yoram Bogacz

  6. Matt Says:

    One of the major arguments made in this two part series is as follows:
 Yoram makes a claim that if evolution by descent with modification were true, then it must be the case that *every single* individual gene can be used to create a phylogenic hierarchy (“Protein or gene comparison is valid, if at all, only if it is consistent regardless of which genes or proteins were used in the comparison”). If that is not the case, he argues, it would “disprove evolution” in some sort of Sherlock Holmes “aha!!” sort of way (that rarely happens in science). He then goes on to present what is called the single-gene discordance problem: that using single genes can produce nonsensical and contradictory results *sometimes*. Thus, he argues, the whole idea of common ancestry falls into question. There are several problems with this line of argumentation, which I will deal with in the following comments.

    
PROBLEM I: The opening premise of this argument is wrong. It is a straw-man argument.

    
A) His own references contradict him:
“To expect that information contained in any single gene will resolve without ambiguity branching events in a taxonomic group with a more than 600 million year history is simply unrealistic.” (Rokas et al, Evolution & Development 5:4 346-359, 2003)
“Here we describe how gene tree discordance can be predicted under a widely used evolutionary model, the coalescent, applied to multiple species.”

    &
    
“One surprising property of gene tree distributions is that the most probable gene tree topology need not match the species tree topology.” (Degnan Rosenberg, 2009)


    B) It also fails to stand up to basic critical reasoning. 
There are obvious reasons why single genes would not produce a useful taxonomy:
    
i) Genetic variability within species: If you and I have different versions of a particular gene and monkey A and monkey B also have different versions from each other, then that gene will not yield a coherent taxonomy. One must choose genes that are common within each species, and it is not always trivial (absent vast interspecies data) to identify such genes.

    ii) Identical genes across species: Humans and chimps have share more than 98% of the same DNA. If I choose a gene that is common to both humans and chimps, I can make no meaningful phylogenetic statements relating us based on that gene alone. An aside: the fraction of shared genes is, in itself, a relatively robust indicator of taxonomic closeness.
    
iii) Genes within a specie continue to mutate (microevolution) after two species diverge: One assumption in Yoram’s “all genes must produce the same branching pattern” argument is that the species remain genetically static after diverging. If they continue to change, this will introduce noise in the system.
These are some of the obvious problems with Yoram’s opening claim. There are many other reasons why single-gene taxonomies are expected to fail. Some of them are more subtle and it is true that not all of them are well understood. But, this is a young field, and the literature shows much progress closing in on the problem.

    
..Which brings me to PROBLEM 2:
Yoram fails to mention that the literature offers many solutions to the problem of single gene discordance problem, and by authors prominently feature in his own bibliography:
Rokas et al (Nature 425, pp798-804, 2003) show that, while single genes yield discordant results, suites of small numbers of randomly chosen genes do in fact produce consistent species trees “with remarkable support.” If Yoram wants to really question genetic relatedness, no number of genes should produce consistent phylogenies. “In this phylogenetic analysis of eight yeast taxa, concatenated data sets of 20 genes are sufficient to provide very strong (>95%) support for the species tree” (ibid). “The results demonstrate that concatenation of a sufficient number of randomly selected genes overwhelms conflicting signals present in different genes” (ibid). Rokas is not the only researcher to demonstrate this. And yeast taxa are by far not the only branch of the tree of life to demonstrate consistent, repeatable genetic trees. In my brief survey of the literature, I have found that the bulk of papers provide simple, well-motivated algorithmic modifications that produce taxonomies that are (a) consistent and (b) agree with other phylogenic approaches. 
In short: Of course single genes do not produce clean taxonomies, but that isn’t the end of the story. Focusing on one problem, while ignoring the explanations, and solutions is not proper scientific skepticism. Focusing *only* on the anomalies, while ignoring the majority of successful phylogenetic reconstructions is likewise improper.

    
PROBLEM #3:
While single-gene trees are difficult to produce, there are many alternative genetic methods to produce reliable and unambiguous demonstrations of inheritance from common ancestors. For me, the most impressive demonstration of common ancestry among the great apes comes from the full genome sequencing of primates and the identification of Endogenous Retrovirus (ERV) insertions. ERV insertions are the remnants of viral DNA that becomes permanently inserted into the genome of the host. These insertions are passed along to all descendants. The likelihood of an ERV insertion to occur at identical point among the billions of base pairs in two unrelated individuals is vanishingly small. For more than one such insertions to coincide is vanishingly small. Humans and chimps share at least 7 retrovirus insertions and this number is expected to increase with further analysis. The number of common ERV insertions among primates seems to correlate strongly with predicted closeness of primate relationships. This sort of coincidence is very difficult to explain by any means other than inheritance. See, for example, Lebedev et al [Gene vol 247(1-2) pp265-77 (2000)].
There are *many* other working methods. Here is a non-exhaustive list:

    1.) phenetic methods
    
2.) cladistic methods

    3.) maximum parsimony
    
4.) maximum likelihood
    
5.) distance methods

    Where there are difficulties with each of these approaches, the combination of these methods has produces consistent and well understood trees across many phyla.

    
To summarize, 
Yes, single-gene methods do not produce consistent trees, but

    1) These are to be expected.

    2) Many of the reasons for this discordance are well understood, though this is a work in progress.
    
3) Many working solutions have been offered.

    Single-gene methods are not the only genetic methods for constructing taxonomies. Many other other genetic methods have proven very successful in producing well-understood taxonomic relationships and produced some unambiguous demonstrations of inheritance from common ancestry.


    Science is not perfect and fraught with places where tidy theories intersect with the messiness of the real world. What a proper informed scientific consumer seeks to do is ask: do multiple lines of evidence using complementary methods robustly demonstrate the principle in question, despite that messiness? To the extent that there are methodological problems and confounding affects (and there will be), can those effects be understood and addressed. Is the science moving towards a better understanding of these challenges or just moving around in circles? Those whose sole interest is in creating doubt and muddying up the water will just pluck the experimental challenges and uncertainties of science out-of-context and string them together in broad rhetorical arguments. They cherry pick quotes and sources only to the extent that those quotes reinforce what they wanted to hear. They look no further and no more rigorously beyond that.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: