Gene Tree Incongruence

(Cornelius_Hunter) #1


Thanks for pointing out Nakhleh, 2013. It is not controversial that gene trees (and the other characters for that matter) are not congruent (i.e., they do not fit the common descent model), and this paper does nothing to dispute that fact. What the paper discusses is progress toward constructing auxiliary mechanisms to “save the theory.”

It is a good example that any Theory A can explain any Evidence B, if you try hard enough. IOW, this paper is written from an evolutionary perspective. That is what they do–try to figure out how to explain the evidence according to evolution. Nothing wrong with that, but we need to understand that this paper is not testing the theory, it is working within the theory.

The argument that CD is a fact because it explains and fits the data so well is no different than the argument that geocentrism or the flat earth are facts. So I’m not citing “remaining challenges,” I’m simply the messenger pointing out that the empirical data do not fit the theory–not from a theory-neutral perspective. It is not even close. I realize this goes against the grain, and in the past I’ve even had people attack me for simply explaining the science. I always forgive them, but the science is the science. It is just non sensical to say CD is a fact (or whatever).

Biological Information and Intelligent Design: Signature in the Ribosome
(Steve Schaffner) #2

No, your statement is quite wrong. It is not controversial that gene trees are not perfectly congruent, but also understood that they are highly congruent, to a degree that is immensely improbable by chance. Since this is true, and since the proposed evolutionary mechanisms for the incongruencies have been observed, and since no competing explanation for the observed pattern has ever been offered, comparative genomics continues to offer powerful evidence for common descent. More powerful all the time, in fact, as more data is collected.

(Benjamin Kirk) #3

That’s not what “not congruent” means.

It’s also predicted that gene trees will not be perfectly congruent, as we know in advance that a mutation and reversion (2 mutations) are not seen and will be misscored as 0 mutations. We swamp out that small systematic error with vast amounts of new data.

(Steve Schaffner) #4

There are multiple processes that lead to inconsistent gene trees: reversion mutation, HGT, incomplete lineage sorting, gene duplication and loss, and mutational saturation. This is a perfectly normal technical challenge of pulling out the cleanest signal from noisy data; the existence of the signal is not in any doubt. The important point is that all of the sources of noise are either directly observed or inevitable given the processes we do observe.

(Dennis Venema) #5

If I had a nickel for every time an incongruent gene tree was used to argue against common descent…

The point is, as has been mentioned, that we expect some (rare) gene trees within a species tree to be incongruent. If we didn’t observe them then we would have a problem.

(Cornelius_Hunter) #6

Oh gosh no. We’re nowhere close to shades of perfection. You are probably just unfamiliar with the literature. MicroRNA and aminoacyl tRNA synthetases are just a couple of examples of types of genes that make no sense on common descent. But this is widespread across gene families (see here for example). It is also widespread in other genetic elements, for example in transcription factor binding sequences. And of course it is also widespread in morphological characters. Even more importantly, and usually not acknowledged, are the cases where you can’t even make a comparison because the gene (or character) does not even exist in all of the species under study. Such characters are simply filtered out, and so not tallied as a failure for common descent. Systematic studies show we’re nowhere close to shades of perfection. The data simply do not fit the common descent model.

Well this is a false dichotomy. IOW, we are not faced with a choice between “chance” and “common descent.” When common descent fails, then evolutionists will draw upon explanations such as common mechanism, convergence, and so forth. Common descent is a specific model that can be tested, and it fails dramatically and repeatedly, thus requiring common mechanism type explanations. “Chance” is irrelevant. Chance is, however, a convenient strawman. It masks the massive failures of the common descent model, and allows us to sustain those failures without concern.

(Cornelius_Hunter) #7

You should have a whole bunch of nickels.

Sure, the common descent model would expect some rare incongruencies, as you say. This is why the model fails–we are nowhere close to some rare exceptions.


Another nickle for your account.

(Dennis Venema) #9

Indeed. Cash or credit? :slight_smile:

(Dennis Venema) #10

Pick any closely related species you like, and run the numbers. I.e. species that have high degrees of shared syntenty, and high overall nucleotide identity. Humans, chimpanzees and gorillas, for example. There are rare exceptions that arise through incomplete lineage sorting, and they are expected, and useful for estimating ancestral population sizes. Trying to claim that this is somehow a problem for evolution is simply failing to understand the evidence.


Bitcoin. Beware that the transaction should take 20-30 minutes to process and may be stolen by hackers.

(Cornelius_Hunter) #12

No, this is not at all normal. Take the transcription factors, for example, A massive rewiring of their binding sites would have had to have occurred. In order to reconcile the data with common descent, enormous amounts of evolutionary change are required.


So I’m reading a paper here about aminoacyl tRNA synthetases…

Gregory P. Fournier, Cheryl P. Andam, Eric J. Alm, J. Peter Gogarten “Molecular Evolution of Aminoacyl tRNA Synthetase Proteins in the Early History of Life” Orig Life Evol Biosph (2011) 41: 621. doi:10.1007/s11084-011-9261-2

(Benjamin Kirk) #14

[quote=“Cornelius_Hunter, post:12, topic:9420”]
Take the transcription factors, for example,[/quote]
OK. All of them or two of them?

[quote] A massive rewiring of their binding sites would have had to have occurred.
[/quote]That seems like a silly metaphor for the subject.

How are binding sites wired, much less rewired, even metaphorically?

(Benjamin Kirk) #15

[quote=“Cornelius_Hunter, post:6, topic:9420”]
Common descent is a specific model that can be tested, and it fails dramatically and repeatedly,…
[/quote]Yet you have this whole series of web pages, and AFAIK you haven’t discussed any of the evidence. You do understand that quotations aren’t evidence, don’t you?

(Steve Schaffner) #16

Who said we were? The question is, do we see more inconsistencies between trees than we would expect from observed and predictable processes? If we don’t, then they are not evidence against common descent – they’re evidence against a process that no one thinks should occur. You’ve presented no evidence that we do, so I’m afraid all of your cases are pointless.

Maybe not, but you sure as heck haven’t made the case that they don’t. [quote=“Cornelius_Hunter, post:6, topic:9420”]
Well this is a false dichotomy. IOW, we are not faced with a choice between “chance” and “common descent.”
Well, chance is the best alternative I can come up with for explaining consistent gene trees, given your failure to provide one.

I wasn’t aware that common descent had failed. Again, making predictions that ignore known mechanisms, as you have done, is not a valid way of testing a theory. I thought you were trained as a scientist – how are you seriously presenting this as an argument? You’re suggesting that there’s something illegitimate about incorporating known physical processes into a scientific model, rather than sticking with a cartoon version.

(Cornelius_Hunter) #17

Well let’s not be tedious. You said “gene trees are not perfectly congruent” [your emphasis].

Yes, that is a good way of putting it. And the answer is “yes.”

Yes, agreed.

Well y’all have made this charge repeatedly even though I’ve provided examples, links, articles, journal papers, and so forth. So I’m curious what kind of broomstick do we have to fetch for you guys? Why, for example, do this and this not qualify as “evidence”?

I can understand people disagreeing on findings. But covering ones eyes and blaming the other guy for not providing any evidence something else altogether. You’ll forgive me if I’m sensing denialism.

(Cornelius_Hunter) #18

Are you serious? That’s quite a justification for a fallacious interpretation of the data. Suddenly it’s all my fault.


There is no known mechanism for completely rewiring transcription factor binding sites, unless “evolution did it” qualifies. Your convenient ignorance of such problems–and there are many more where that came from–allows great confidence.

(Steve Schaffner) #19

Great! Where have you published your calculations?

(Steve Schaffner) #20

Evidence of what? If the first article is right, and microRNA content is highly consistent across species, then they give conflicting answers to other sources about the deepest mammalian branches – precisely the places where the signal is most difficult to make out already. If the second article, then microRNA content is not highly consistent across species, and we should ignore the conclusions of the first paper. Which is it?

Again, denialism of what? I’m perfectly happy to compare the strengths of competing explanations, but you’re not offering an explanation for anything.