Elsewhere on the internet...

The League of Reason has some social media accounts! You can find us on Facebook or on Twitter for some interesting links and things.

Debate: dandan vs Inferno

POSTING RESTRICTED

Posting in this topic has been restricted to the following users:

Inferno, dandan.
Post new topic   Page 1 of 2
 [ 26 posts ] 
Debate: dandan vs Inferno
Author Message
ProlescumWebhamsterUser avatarPosts: 5002Joined: Thu Dec 31, 2009 8:41 pmLocation: Peptone-upon-Sores

Post Debate: dandan vs Inferno

This thread is for the debate between dandan and Inferno on the subject:




Do phylogenetic relationships hold up to the evidence?






Each combatant will have ten posts to make their case, at which point, the thread will be closed and the victorious argument decided by Isotelus (see discussion thread link below).

As I understand it, there is no word limit (Gish gallops will likely be mocked, however) and no time limit between posts.



A specific thread for members to discuss this thread is available here.

Neither participant may post in the discussion thread until the debate is complete.

First up, I believe, is Inferno.
if constructive debate is allowed to progress, better ideas will ultimately supplant worse ideas.

Comment is free, but facts are sacred
Wed Jun 25, 2014 9:12 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

Prolescum wrote:This thread is for the debate between dandan and Inferno on the subject:




Do phylogenetic relationships hold up to the evidence?






Each combatant will have ten posts to make their case, at which point, the thread will be closed and the victorious argument decided by Isotelus (see discussion thread link below).

As I understand it, there is no word limit (Gish gallops will likely be mocked, however) and no time limit between posts.



A specific thread for members to discuss this thread is available here.

Neither participant may post in the discussion thread until the debate is complete.

First up, I believe, is Inferno.

AGREE
Wed Jun 25, 2014 9:55 pm
ProlescumWebhamsterUser avatarPosts: 5002Joined: Thu Dec 31, 2009 8:41 pmLocation: Peptone-upon-Sores

Post Re: Debate: dandan vs Inferno

That post won't count toward your overall allowance. Please observe the formalities of the debate forum ;)
if constructive debate is allowed to progress, better ideas will ultimately supplant worse ideas.

Comment is free, but facts are sacred
Wed Jun 25, 2014 10:02 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

Thank you Prolescum for setting up the debate thread, thank you Dandan for the opportunity.

The topic is "Do phylogenetic relationships hold up to the evidence?" and I will be arguing the positive.
In a first step, I will discuss very briefly the history of phylogenetic classification. I will brush on Carl von Linné and his system of classification and briefly explain why it is inadequate to explain the relationships between living things.
I will then, after a very brief detour into the philosophy of science, formulate a hypothesis and weigh it against the evidence. This will include Open Access papers, but also the odd quote from University-level biology textbooks.
Finally, I will explain why creationist systems of classification don't hold up to the evidence. This last part isn't strictly needed and I won't fault Dandan if he ignores it, he has every right, but it might prove an interesting talking point.

Note that I will use "organisms" instead of saying "animals, plants, etc.". The use of "organisms" is supposed to refer to all life as we know it.
Also note that I will use "unrelated" to talk about very distantly related organisms. For the purpose of this post, for example a human and yeast are "unrelated" while a human and a monkey are closely related.

Further note that I am a layman and will probably make a few (a lot of) mistakes. I expect that the peanut gallery will explode in a frenzy of corrections after I have posted this. However, I won't note those until this debate is over, which might be quite some time.

----------------------------------------------------------------------------------------------------------

Phylogenetic classification - A brief history

The purpose of any system of classification is to group a variety of things, either natural or produced by humans, living or dead, fictional or real, etc. into schemata that allow easy comprehension of the things at hand. For example, one might group pencils by their size or by their hardness or by both. The best system is the one that best copes with this variety.

Living things can be classified according to artificial or natural parameters. The most important system used during the last three-hundred years used to be the one constructed by Carl Linnaeus (or Carl von Linné). He classified living things according to artificial parameters: For example, plants were classified according to how many stamen a plant had. This system might be practical when classifying or identifying plants, but Linné realized that other systems could be more natural and reflect their genealogy or ancestry: This system would be called "phylogenetic system".

Living things, as we know, have ancestors. I have a mother and a father, each of whom again have a mother and a father. It seems logical to construct a classification of life by grouping closely related organisms together, while more distantly related organisms are further away in the grouping. This is reflected in family trees: My brother looks nothing like my father nor like me, he looks more like his uncle. Both are rather short and brownish-blonde, while both my father and I are tall, handsome and have dark-brown hair. An artificial system of classification might group my brother and uncle together, with my father and myself in a separate group. However, my brother is more closely related to my father and me, in that we share more genes in common with him than does my uncle. That's why our family tree groups us more closely together rather than the way an artificial system might.

This second, natural system is also a very objective one. We might be undecided whether "pencil length" or "pencil hardness" is the characteristic trait, but we're absolutely sure that my brother is my fathers son. There's absolutely no doubt about it.

The first, rather rough attempts at classification were done through comparative anatomy. Similar organs or structures in different organisms were grouped together, divergent organs and structures were not. Identical structural elements were called "homologous". Organisms with no structures in common were grouped far apart, organisms with a few homologous structures were grouped closer together, organisms with many homologous structures were grouped even more closely.

For example, we'll look at a human being, a dog and an ant. Specifically, we'll look at their hands/arms and the bones in them. Both the dog and the human have radius and ulna in their arms, while the ant has no bones. According to homology, we'd group humans and dogs together while the ant is the odd one out.

Let's shake things up a bit: We'll look at the arms again, this time of a human, a dog, a bird and a whale.

Image

All four animals have a radius, an ulna and carpal bones. Without further study, we'd not get very far in classifying them. For the purpose of this discussion, we can't go any further. (We could, of course, but I want to come to an end at some point. Suffice it to say that comparative anatomy, which is what this is, does have its limits. We now need something more than that.

I could, at this point, explain the unity of life, nested hierarchies and fossils, but I want to delve into the genetic line of evidence. I might go into them if further posts require it.


Phylogenetic classification - Philosophy of science, hypothesis construction and the evidence

Richard P. Feynman wrote:"... there are many reasons why you might not understand [an explanation of a scientific theory] ... Finally, there is this possibility: after I tell you something, you just can't believe it. You can't accept it. You don't like it. A little screen comes down and you don't listen anymore. I'm going to describe to you how Nature is - and if you don't like it, that's going to get in the way of your understanding it. It's a problem that [scientists] have learned to deal with: They've learned to realize that whether they like a theory or they don't like a theory is not the essential question. Rather, it is whether or not the theory gives predictions that agree with experiment. It is not a question of whether a theory is philosophically delightful, or easy to understand, or perfectly reasonable from the point of view of common sense. [A scientific theory] describes Nature as absurd from the point of view of common sense. And it agrees fully with experiment. So I hope you can accept Nature as She is - absurd.

I'm going to have fun telling you about this absurdity, because I find it delightful. Please don't turn yourself off because you can't believe Nature is so strange. Just hear me all out, and I hope you'll be as delighted as I am when we're through."

-Feynman, R. P. (1985) QED: The Strange Theory of Light and Matter. Princeton, NJ: Princeton University Press.


When we now come to the hypothesis and the evidence, I want you to bear the above in mind. No matter how absurd something might sound to you, if it agrees with the available evidence then it must be accepted as the current best explanation.

In 1964, Richard Feynman gave his famous lectures as part of the also-famous "Messenger Lectures". In the seventh lecture, he talks about how to find new laws. Of particular interest is the part between 16:40-18:24. In these few minutes, he presents what is in its essence the scientific method. I will try and hold myself to his suggestions, though I might have some difficulty with the second step.

His three steps are:
1) Take a guess.
2) Compute the consequences of the guess.
3) Compare the results with the evidence found in nature.

The guess I take is the following:
Relationships between organisms are reflected by their shared ancestry.

The more difficult bit is how we can compute the consequences. Let's take a step back first. Above, I said that there's no doubt that my brother is indeed my fathers son. This is only true if we're certain that my mother wasn't inseminated by someone else. But there's another way as well: A paternity test. By comparing several stretches of DNA, one can almost conclusively say that some in or is not a person's father. I say "almost conclusively" because these tests are statistical, so do not provide absolute truth and because so-called chimeras exist, which produce false negative results. Typically, paternity tests (and other DNA tests) produce results with an accuracy of over 99%. These paternity tests are used in courts of law, so I will assume that they are indeed accurate.

These paternity tests could also be used on other organisms, right? We could use genetic markers to figure out how closely related two (or more) organisms are.

I'll skip the exact mechanisms as I suppose that they're known. Basically, genetic markers (sequences of DNA) are chosen according to a few criteria and are then compared using a variety of methods, in the case we're talking about they used MCMC-based Bayesian inference.

So we'll move to the second step: Computing the consequences. What should we find if DNA-testing can indeed reveal the ancestry of organisms and are organisms really related?
The latter is easy: We should find a very small number of similar trees with extremely large likelihoods, while the millions and billions of other trees are very unlikely indeed. We expect to find some discrepancies, either due to known factors (Homopalsy, horizontal gene transfer, etc.) or because all the data is not in and new data constantly updates the models.
We would also expect that obviously unrelated organisms such as humans and yeast remain very far apart on the trees, while closely related organisms are also closely related according to these models. For example, it would very much surprise us if humans were more closely related to cats than they are to monkeys. If we would find the statistical likelihood pointing in that direction, evolution would be in trouble.

The former is slightly more difficult. As explained above, we do expect some slight discrepancies. We expect that morphological evidence is fairly robust, but that genetic evidence will update some smaller errors. If morphological trees would be completely different from genetic ones, we might be in trouble. For example, morphology reveals that humans are more closely related to dogs than they are to ants, so we'd expect genetic evidence to confirm that. (I'm using extreme examples to make a point.)

I'll try to explain that a little better: Humans and dogs definitely share at least one common trait, but we'll focus on ulna and radius just now. If organisms without bones were classified more closely with one of the two to the exclusion of the other, in other words a ((H,A)D) tree, that would utterly shatter phylogenetics. That's what I mean when I'm talking about classifying distantly related with closely related organisms.

However, morphological data might be insufficient to pinpoint the precise relationships between humans, chimps, orang-utans and gorillas. That's why we need genetic information to draw up the correct trees. In that case, it's (((H,C)G)O).

Just in case this is unclear, I'll put an explanation below the fold:


Another problem might be that we could be related to other organisms at equal distance, that is to say that (((H,C)G)O) and (((O,H)C)G) as well as all other possible ways to form that tree would be of equal probability. They're not and I'll be happy to go into that if you want to seriously challenge that.

Now as an in-between, we need to know if cladistic analysis actually works. Will these analyses find the correct tree? As previously stated, they are already allowed in a court of law, but that's not good enough for me. What if we took some organism with a short generation span and tracked their lineages, that is to say we know exactly who the offspring of whom is. Out of over 130,000 possible trees, will cladistic analysis find the correct tree? As it turns out, yes it will.
TalkOrigins lists a few of these examples, so we know that cladistic analysis is damn accurate.


Out of all the possible examples, I'll pick one particular clade and go from there.
In an earlier post, Isotelus corrected Radio-host Bob Enyart on the relationship between bats, cows and horses. For that post, I adapted the cladogram from Zhou et al. (2012) to highlight the various relationships in the superorder Laurasiatheria. (Page 156, figure b.)

Image

I already explained above (below the fold) that some discrepancies will occur and that they did indeed occur. However, cladistic analysis correctly placed humans and mice as outliers, while other groups were placed closely together. I made suggestions on how phylogenetics could be falsified, but neither one proved true: More distantly related organisms weren't closely grouped, the reverse was true. Different trees were also supported with different levels of support, the best ones (figure b, or the picture above) were supported with maximal BP support (100%).

I'll try to make that point absolutely clear: There are 15 different taxa in the cladogram I posted, the tree itself is a rooted tree. So for NR = (2n-3)!! = (2n-3)!/(2n-2(n-2)!), we can expect 213,458,046,676,875 different trees to be constructed. However, this one was found with maximal support. Imagine the sheer unlikelihood for that!


Phylogenetics - The creationist hypothesis

As I explained above, and I hope I'm doing a good job in my sleep-deprived state, organisms are definitely related to one another, with methods even accepted in court.

If creation is true, however, we would expect that animals are indeed related up to a certain point (for example, not even creationists dispute that all dogs are related) but that this relationship breaks down at some point. So if there were some credit to creationism, we might expect that cladistic analysis shows no relationship between cats and horses or humans and dogs or whatever category you want to draw up. If that were the case, we'd expect the relationships to be at equal distance to one another, thus making hundreds of trees equally likely.

However, this is decidedly not so. As Zhou et al. (2012) clearly shows, there is a definite relationship between animals that creationists would consider to be different kinds.

I already explained in a different post that creationist research methodology doesn't conform even to elementary university standards, but I also showed that the inferences they make are purely subjective and that they dismiss the more rigorous forms of evidence.

But I'll trump even that. In Feynman's speech I noted earlier, between 26:14-29:06, he notes that a new guess (or hypothesis) must be better than the earlier hypothesis or theory. There are some guesses that are obviously wrong, so we needn't even consider them for any subsequent model.

For any creationist model to work, we'd need either a definition of what kinds are or at least a definite model on how to distinguish between kinds. In other words, how can you be sure that the relationship between organisms stops? AronRa put that very well in his Phylogeny Challenge:
Inferno wrote:Are mallards related to pochards, wood ducks and muscovys? Are all ducks also related to geese and other Anseriformes? Are Anseriformes related to Galliformes and other Neognathae? Are Neognathae related to Palaeognathae? Are any extant (still alive) birds related to Hesperornis, Ichthyornis, Enantiornis or other Euronithes? Are Euronithes related to Confuciusornis or Archaeopteryx? Are all early Aves (birds) related to Microraptor, Velociraptor or other non-avian Dinosaurs? Are dinosaurs related to Pterosaurs, Phytosaurs and other Archosaurs?


Image

If creationism is correct, then some of these animals are related, but they are unrelated to everything else. Can you pinpoint where these differences occur? Which of these are related? Which of these are created?

There was once the hope that baraminology might save creationism. Creationists suggested different methods of identifying kinds and showing how they are (un)-related. A few of those ways I criticized in one of the posts above.
Creationists are typically not taken seriously by scientists, but one scientist did: In 2010, P. Senter published an article in the Journal of Evolutionary Biology. In it, he used creationist methods to determine how (un)-related animals were. He found, unsurprisingly to scientists, that groups classified by creationists as distinct "kinds" were related even according to the models used by creationists.



Conclusion

I hope that I was able to show that morphological analysis on its own is insufficient to establish correct evolutionary relationships, so genetic analysis and evidence is needed. Genetic evidence is extremely robust and produces results which closely match the predictions previously established, but which also shed new light on relationships that could not be properly established by morphological analysis alone.

I hope that I was further able to show that the hypothesis, that at the very least all Boreoeutheria are related, has not yet been disproved. Phylogenetic relationships established by recent DNA testing proved to conform rather well with earlier predictions and establish, beyond reasonable doubt, that life is indeed related.
Finding the correct tree is incredibly unlikely, yet both blind tests and ongoing research hones in on the correct trees with astonishing probability.

Third, I hope that I was able to show that creationist hypothesis and theories fail all tests they are put to. Any new hypothesis or theory must supplant the old one and in that, creationism utterly fails. Phylogenetic relationships, and subsequently common ancestry, are the currently most robust systems of classification we have. Creationists own methods have been used against them to show that their methods are bogus. Any attempt to overthrow the current model must have better predictive capabilities.

Zhou, X., Xu, S., Xu, J., Chen, B., Zhou, K., & Yang, G. (2012). Phylogenomic analysis resolves the interordinal relationships and rapid diversification of the laurasiatherian mammals. Systematic Biology, 61(1), 150–64. doi:10.1093/sysbio/syr089

Senter, P. (2010). Using creation science to demonstrate evolution: application of a creationist method for visualizing gaps in the fossil record to a phylogenetic study of coelurosaurian dinosaurs. Journal of Evolutionary Biology, 23(8), 1732–43. doi:10.1111/j.1420-9101.2010.02039.x
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Thu Jun 26, 2014 11:15 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

[sThanks for this opportunity, I will proceed with my first reply, but I what to make clear that I will be out of town next week, so I will probably post my second reply and the end of next week.

I'll try to make that point absolutely clear: There are 15 different taxa in the cladogram I posted, the tree itself is a rooted tree. So for NR = (2n-3)!! = (2n-3)!/(2n-2(n-2)!), we can expect 213,458,046,676,875 different trees to be constructed. However, this one was found with maximal support. Imagine the sheer unlikelihood for that!


There is no surprise there, if any other tree would have predominated you would have simply said that the other three is the correct tree. For example if humans would have had feathers and beaks, instead of nipples and placentas, you would have simply classify humans as “birds” and you would be saying that this is the “correct” tree.

If some mammals would have had feathers you would simply conclude that feathers evolved before birds and mammals diverged, and you would call that the “correct tree” for example some birds have teeth (like mammals) and you explain this by stating that teeth evolved before birds diverged from mammals.

The point that I am trying to make is that there is nothing special about our current family tree, if things would have been different, for example if mammals would have had feathers, or if birds would have had placentas, you would have simply drawn a different family tree, and you would be calling that the “correct tree.” One tree has to predominate no matter what, this would be true regardless if life evolved, or if life was design, or if life was created by some other mechanism.

Besides, evolution didn´t predict that mammals have nipples and birds have feathers, everybody knew that since long before Darwin.
However I understand how evolution is supposed to work and I understand why a PERFECT nested hierarchy could only be reasonable explained with common descent, ¿why would a designer create life in order to fit a perfect nested hierarchy as if evolution(common ancestry) where true?
However as we both know we don´t have a perfect nested hierarchy, but rather we do find some discordances, as you admited
We expect to find some discrepancies, either due to known factors (Homopalsy, horizontal gene transfer, etc.) or because all the data is not in and new data constantly updates the models.

So the real question is, are this discordances statistically significant?

I will point to some discordances, however I what to make clear that I am presenting these as examples, my point is that discordant trees are pretty much the rule and not “strange exceptions”

Many of the first studies to examine the conflicting
signal of different genes have found considerable
discordance
across gene trees: studies of hominids [9–11],
pines [12], cichlids [13], finches [14], grasshoppers [15] and
fruit flies [16] have all detected genealogical discordance
http://web.stanford.edu/group/rosenberg ... 9-TREE.pdf


In 30% of the genome, gorilla is closer to human or chimpanzee than the latter are to each other;
http://www.nature.com/nature/journal/v4 ... 10842.html


The genes (that code for the platyuss venom) resemble those of other venomous animals, such as snakes, lizards, starfish and sea anemones
http://www.nature.com/news/2010/101012/ ... 0.534.html


Echolocation is a sensory mechanism for locating, ranging and identifying objects which involves the emission of calls into the environment and listening to the echoes returning from objects [1] . Only microbats and toothed whales have acquired sophisticated echolocation, indispensable for their orientation and foraging [1] . Although the bat and whale biosonars originated independently and differ substantially in many aspects [2] , we here report the surprising finding that the bottlenose dolphin, a toothed whale, is clustered with microbats in the gene tree constructed using protein sequences encoded by the hearing gene Prestinhttp://www.cell.com/current-biology/abstract/S0960-9822(09)02057-0


We identified a human endogenous retrovirus K (HERV-K) provirus that is present at the orthologous position in the gorilla and chimpanzee genomes, but not in the human genome
http://www.ncbi.nlm.nih.gov/pubmed/11378389


I am obviously not expecting you to review each of this articles, and I encourage you not to, unless you consider it relevant, but my question is ¿how do you know that these discordances are not statistically significant? What objective method can you provide to determine if a given discordance is significant?

Most genomes have never been sequenced in detail, which means the more discordances are likely to be discovered as new genomes are sequence. ¿what objective method can you provide so that we can know in advance which potential sequences would provide a problem for evolution?

Statistical significance is a well understood concept and there are equations (http://en.wikipedia.org/wiki/Statistical_significance) for example if I argue that Germans are taller than Mexicans and I find one Mexican that is taller than an other German, there are objective ways to show if this discordance is statistically significant.

In other words, under what objective bases would you say that the genetic homology between bats and whales in echolocation is statistically insignificant? (I what to see the numbers)


1 ether we find a perfect nested hierarchy

2 orwe have statistically insignifican discordances

3 or have statistically significant discordances
The creationists model is only consistent with the third option, this is why it is very important to provide a statistical model
Fri Jun 27, 2014 4:39 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

You have all the time you need to write your reply, there's no need to hurry.

Now I have to admit that I'm a bit peeved: I wrote what I thought was a fairly good attempt at explaining why phylogenetics is robust, but you take exactly one quote out of the whole thing and look at only that. Very well, why not. I'll try to address your every point.

----------------------------------------------------------------------------------------------------------

dandan wrote:
Inferno wrote:I'll try to make that point absolutely clear: There are 15 different taxa in the cladogram I posted, the tree itself is a rooted tree. So for NR = (2n-3)!! = (2n-3)!/(2n-2(n-2)!), we can expect 213,458,046,676,875 different trees to be constructed. However, this one was found with maximal support. Imagine the sheer unlikelihood for that!



There is no surprise there, if any other tree would have predominated you would have simply said that the other three is the correct tree. For example if humans would have had feathers and beaks, instead of nipples and placentas, you would have simply classify humans as “birds” and you would be saying that this is the “correct” tree.

If some mammals would have had feathers you would simply conclude that feathers evolved before birds and mammals diverged, and you would call that the “correct tree” for example some birds have teeth (like mammals) and you explain this by stating that teeth evolved before birds diverged from mammals.

The point that I am trying to make is that there is nothing special about our current family tree, if things would have been different, for example if mammals would have had feathers, or if birds would have had placentas, you would have simply drawn a different family tree, and you would be calling that the “correct tree.” One tree has to predominate no matter what, this would be true regardless if life evolved, or if life was design, or if life was created by some other mechanism.


There are three replies to be made here and I will formulate them one after the other.

Different life - different tree

Your first argument is that had life evolved differently, we would expect different phylogenies. I'm not exactly sure how that is an argument against anything, but I'll try and address it nonetheless.

Imagine ordinary family X: The Simpsons.

Image

Homer and Marge married and had the three lovely kids we know from the series. But what if Homer had married Patty instead? Why, the whole tree would look different. Their children would have different characteristics and would not look like the Bart, Lisa and Maggie we know today. If that were the case, we'd obviously have to draw the tree in another manner: Marge and Patty would swap places, Bart/Lisa/Maggie would not exist and in their places, other children might be drawn.

Life either evolved exactly one way or it didn't evolve at all. There is little point in asking "what if life evolved differently?" because it can't inform on real life. You're asking Sheldon Cooper's question "Have you ever wondered how humans would be different if they evolved from lizards instead of mammals?" Some people think about that stuff, but why?

There are millions of species around today and we're trying to see how they relate to each other. Now if scientists had a priori decided that we had evolved from monkeys and then made the evidence fit that hypothesis, then your question might make some sense. But it's the other way around: Nobody wants to be related to a monkey, it's just the way things are.

I'll go back to our Simpsons family tree. If I looked at only the family members without the tree, I might conclude that Homer and Herb are related, because they share several traits in common. I might note that Jackie is Marge's mother, that Patty and Selma are related and that Lisa and Maggie are related. All of them share several traits in common. I'll disregard Ling because she's an adoptive daughter, it would needlessly complicate the analogy.

So I've got those groups together, but I'd be at a loss as to where they fit. I can't let guesses cloud my judgement, so I'd perform paternity tests. My first surprise would be that Herb is only Homer's half-brother. They share so many traits in common, I'd have been surprised at that. But nature doesn't care how surprised I am, she simply is. And on it goes: I might not be terribly surprised that Marge and the twins are related, but I might be surprised that the only surviving lineage is that of Homer/Marge.

In almost exactly the same way, I described the sorting of Laurasiatheria. The mere fact that we're constantly surprised by new discoveries (at some point it was thought we were more closely related to gorillas than to chimps; then there's the whole bit about Xenarthra I explained; and so on) puts your argument to rest. There are no a priori decisions as to what is related to what, we simply follow the evidence.

The idea would only be valid if there were a huge conspiracy among scientists. If you think that is the case, then we can cease this discussion right now because the evidence would be tainted anyway.

Finding the correct tree

Your second argument might be construed as suggesting that the methods used aren't solid enough or that all trees are equally probable and that scientists simply picked out the tree they wanted. This is decidedly not so and it's trivial to prove it. I'll be brief on that.

I explained above that scientists actually put phylogenetic analysis to the test: They watched organisms evolve in real time and took note of their lineages. That is to say they constructed a family tree in real time. They already knew what the real tree would look like.
They then took some of the genetic markers from the extant lineages and input them into five different phylogenetic models. If phylogenetic analysis were incorrect, if the methods weren't solid enough or if there were some other problem, we might expect all lineages to be equally likely or we might expect some other problem. But if phylogenetic analysis is correct, then we'd expect either the absolute correct tree or we'd expect two or three trees that match the real one extremely well. (I already explained that small discrepancies are to be expected, especially in organisms where horizontal gene transfer can occur easily.
As I also explained, the five models all brought back the one correct tree.

Inferno wrote:Now as an in-between, we need to know if cladistic analysis actually works. Will these analyses find the correct tree? As previously stated, they are already allowed in a court of law, but that's not good enough for me. What if we took some organism with a short generation span and tracked their lineages, that is to say we know exactly who the offspring of whom is. Out of over 130,000 possible trees, will cladistic analysis find the correct tree? As it turns out, yes it will.
TalkOrigins lists a few of these examples, so we know that cladistic analysis is damn accurate.


This was a direct test of cladistic analysis and it passed with flying colours. It found the correct tree even though the model didn't know what the correct tree was: It had to test out the relationships and construct a tree according to the likelihood of them being related in such a way. Phylogenetic analysis undoubtedly works in blinded tests in the laboratory, so we'd expect them to work in the "real world" too, right?

Mammals with feathers

Your third argument is a bit of a rehash of the first argument, but I want to quickly address it nonetheless. What would happen if mammals with feathers were found? As far as I know, but I might be wrong on that, feathers have only been found in theropod dinosaurs and their descendants: Birds. If their exact feathers, with the same structural development and genetic coding were found on one mammal, but not on any other mammal, then we'd have a serious problem. I'd go as far as to say: That would utterly disprove evolution.

What is not uncommon, however, is convergent evolution. Maybe, in a few million years, the hair on bats and/or flying colugos might evolve to become something closely resembling avian feathers, even if they're structurally completely different. That wouldn't phase any scientist in the slightest, it would simply mean that evolution is taking place.



dandan wrote:However I understand how evolution is supposed to work and I understand why a PERFECT nested hierarchy could only be reasonable explained with common descent, ¿why would a designer create life in order to fit a perfect nested hierarchy as if evolution(common ancestry) where true?
However as we both know we don´t have a perfect nested hierarchy, but rather we do find some discordances, as you admited
Inferno wrote:We expect to find some discrepancies, either due to known factors (Homopalsy, horizontal gene transfer, etc.) or because all the data is not in and new data constantly updates the models.


I doubt we'll ever find the one, perfect tree of life for all organisms. We'd need the information of all extant species, we'd need perfect sorting models and we'd probably need a good deal of genetic information from extinct animals, something we can't hope to get. (At least not in these quantities) An absolutely perfect tree of life is out of the question. Remember that we've only known about DNA for a little under 150 years, that the double helix shape was only suggested 61 years ago and that the human genome was sequenced a mere eleven years ago. If such a perfect model is theoretically possible, and I don't think it is, then we have a few hundred years of research before us before we can actually find it.

Instead, we can expect ever better trees, with new evidence constantly updating the old ones. What we need to accept that our current trees are extremely close approximations is direct evidence that phylogenetic analysis works, such as the one I provided above, as well as robust evidence that our current trees are backed up by multiple lines of evidence. I did both of those.

----------------------------------------------------------------------------------------------------------

We now come to the five examples that supposedly show problems with phylogenetics. I'll quickly explain why none of them are any problem at all, though I won't necessarily go into individual papers. I'll also point you to the explanations above as to why discrepancies are to be expected. Note that scientists know why they occur and how they can deal with them.

I'll number the five examples 1-5 in order, just so we know what I'm talking about.

Number three already has the answer in the line you quote: "resemble". This means "similar, but not the same". I didn't even have to open the nature article or read the study to know that they were talking about convergent evolution. Turns out I was right: "Poisonous platypuses confirm convergent evolution"
The paper mentions that many of the proteins are functional analogues, meaning that they are structurally different but have similar functions.

The paper goes on to note that the venoms evolved independently:
"The fact that these extremely divergent species share similar venom components, some of which were found repeatedly in platypus and other venoms, suggests that there are indeed protein motifs that are preferentially selected for independent evolution to venom molecules in a striking display of convergent evolution, and that many animal venoms share some similarities in their mode of action [27]."

Numbers one and two have already been dealt with by both Isotelus and MasterGhostKnight. I won't go into that except to point out that it's been done:
Here, here, here, here and here. I simply won't waste my time dealing with individual papers.

I will, however, take the time to link you to the following blog post over at BioLogos. I'm not particularly fond of them, but the post is excellent.
It explains why gene trees are not always exactly the same as the species trees and it's the species trees we're concerned with. (That directly fouls papers 1, 3, 4 and 5) This also explains why the rest of your argument is invalid: You're looking for discordant gene trees, but they're to be expected. That's why you look at multiple loci instead of just one and that's also why you look at multiple lines of evidence.

As Isotelus pointed out, "[d]iscordance would falsify evolution if studies showed both consistent and significant mismatches between gene and species trees". You're showing individual discordant genes, but even they fit into the overall lines established. That's what I showed when I provided the Laurasiatheria paper: Multiple lines of evidence with scientists looking at multiple genes provide the evidence, singular genes might not reflect that due to incomplete lineage sorting.

dandan wrote:I am obviously not expecting you to review each of this articles, and I encourage you not to, unless you consider it relevant, but my question is ¿how do you know that these discordances are not statistically significant? What objective method can you provide to determine if a given discordance is significant?


I'm not a statistician, in fact I'm particularly poor at it, so I'd be hard-pressed to come up with any statistical model, no matter what it's for.

However, I'd again refer you to papers where multiple loci are analysed.
I'll make this absolutely clear: Individual genes being discordant don't pose a challenge to anything at all. It is the wealth of loci that provide the evidence. If analyses which look at multiple loci continuously disagreed, that would indeed pose a problem. I have not yet seen such papers.

dandan wrote:Most genomes have never been sequenced in detail, which means the more discordances are likely to be discovered as new genomes are sequence. ¿what objective method can you provide so that we can know in advance which potential sequences would provide a problem for evolution?


I actually explained that above:
Inferno wrote:What would happen if mammals with feathers were found? As far as I know, but I might be wrong on that, feathers have only been found in theropod dinosaurs and their descendants: Birds. If their exact feathers, with the same structural development and genetic coding were found on one mammal, but not on any other mammal, then we'd have a serious problem. I'd go as far as to say: That would utterly disprove evolution.


I also explained that multiple trees with equal probability might pose a problem or organisms that could not possibly have been derived from their ancestor (cats sorted into plant lineages or birds into spiders) would indeed pose problems for evolution. As I said, the three organisms "human", "dog" and "ant" could be grouped a few ways, but consistently we see a ((H, D)A) grouping, not any other one. Why is that? I'll tell you why: Because phylogenetic sorting is statistically robust.

dandan wrote:In other words, under what objective bases would you say that the genetic homology between bats and whales in echolocation is statistically insignificant? (I what to see the numbers)


I'm not at all sure that bats' and whales' echolocation is due to a shared ancestry of the genes. In fact, that would be hugely surprising. At most, I'd expect them to share very underlying genes, but certainly not a majority. As was shown in the cladogram in my first post, Chiroptera and Cetacea are fairly far apart, with none of the other Scrotifera being able to use echolocation. I'd instead propose another occurrence of convergent evolution. In fact, that's what the authors of this paper found.

As I said, statistics is one of the few areas of science where I have to shamefully hang my head and declare my absolute ignorance. I simply don't know. All I can do is point to the analysis of other scientists and show what they have done:
Ibid. wrote:Two parallel mutations in Pcdh15 (I946M and E1278D) were shared by branches b, d, and g. Parallel evolution was statistically significant (P<0.001) between branches b and g for the following 22 parallel mutations: N218D, Q310E, E393V, T427S, A433V, I438V, T490I, V546F, I643V, N666K, K820R, I853V, K856T, M946I, V952A, R999L, T1139R, F1160L, A1173S, K1275R, D1278E, and I1404V. Parallel evolution was also statistically significant (P<0.001) between branches d and g for three parallel mutations: A726D, M946I, and D1278E. Finally, eight parallel mutations occurred between branches b and d, including L423V, Q468P, H765Y, F876L, M946I, F984S, V1019I, and D1278E (Figure 2C). Parallel evolution between these two branches was statistically significant (P<0.001). The positions of these sites in the domain structure of Pcdh15 were mapped in Figure S4. The BI tree based on amino acids excluding all parallel-evolved sites differed somewhat with the species tree, but it did not group echolocators together (Figure S5).


I have to assume that the authors used a standard p-value test, so a value of p>0.1 would confirm your hypothesis, whatever that might be. That they do share genes for echolocation? I'm not sure I understand your position, but I understand enough to know that it's wrong.

dandan wrote:1 ether we find a perfect nested hierarchy

2 orwe have statistically insignifican discordances

3 or have statistically significant discordances
The creationists model is only consistent with the third option, this is why it is very important to provide a statistical model


If I understand you correctly, and I'm not sure I do, then number one is practically impossible and number three is wrong, leaving us with number two.



----------------------------------------------------------------------------------------------------------

As an aside, I'd invite anyone with an interest in phylogenetics to watch AronRa's excellent "Falsifying Phylogeny" series and his superb "Turns out we DID come from monkeys!". All seven! videos provide an insight into many aspects of phylogenetics and they actually provide means of falsifying it.

Dandan: Your best hope is in finding an instance that would actually falsify common ancestry, like in the 6th video of the "Falsifying Phylogeny" video. AronRa gives multiple examples as to what would actually falsify it, so why don't you give me one of those?

I'll repeat what I said:
Inferno wrote:For any creationist model to work, we'd need either a definition of what kinds are or at least a definite model on how to distinguish between kinds. In other words, how can you be sure that the relationship between organisms stops? AronRa put that very well in his Phylogeny Challenge:
Inferno wrote:Are mallards related to pochards, wood ducks and muscovys? Are all ducks also related to geese and other Anseriformes? Are Anseriformes related to Galliformes and other Neognathae? Are Neognathae related to Palaeognathae? Are any extant (still alive) birds related to Hesperornis, Ichthyornis, Enantiornis or other Euronithes? Are Euronithes related to Confuciusornis or Archaeopteryx? Are all early Aves (birds) related to Microraptor, Velociraptor or other non-avian Dinosaurs? Are dinosaurs related to Pterosaurs, Phytosaurs and other Archosaurs?


Image

If creationism is correct, then some of these animals are related, but they are unrelated to everything else. Can you pinpoint where these differences occur? Which of these are related? Which of these are created?
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Sat Jun 28, 2014 2:55 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

Ok, I will make some statements and please tell me if you agree with them, this will help me to identify the exact points of disagreement.

1) Some tree has to predominate regardless if evolution, creation or some other model is true.

With “tree that predominates” I mean that when comparing the genome in different organisms, some of them would necessary be more similar to others.

Agree? If not please tell me specifically why do you disagree.



2) If for example orangutans would have been more similar to humans than chimps, you would have drawn a different tree with humans and orangutans in the same clade, and you would have put chimps in a “distant clade” and you would have call that the correct tree. That wouldn´t falsify common descent, it would simply chance the order in which different organisms diverged.

Agree? If not please tell me specifically why do you disagree.



3) A concordant nested hierarchy, with statistically insignificant discordances would prove common descent with near 100% certainty, and ID and other alternative models would be wrong or latest inferior to common descent.

Agree? If not please tell me specifically why do you disagree.


4) A discordant nested hierarchy, with statistically significant discordances would falsify common descent and ID would be consistent with the data. In other words in a ID model any attempt to organize organisms in a Nested Hierarchy would be full of discordances.
Just to be clear A discordant tree is when a portion of DNA does not match the predominant tree, for example if the predominant tree puts humans more closer to chimps than to mice, a discordant gene would be a gene that puts humans closer to mice than to chimps.

Agree? If not please tell me specifically why do you disagree.


So if you if you agree with all these points, then the only point of disagreement is if the discordances that we observe are statistically significant.
What objective criteria are you using to determine that a mammal with feathers would falsify common decent, but something like a platypuss with “snake venom” won´t falsify it? What is fundamentally different between these 2 examples?

Under what objective criteria do you say that 30% discordances in the human/Chimp/Gorilla genome is insignificant? What if instead of 30% we would have had 35% or 40% or 50% up to what point would you say that these discordances are significant?

In other words, what is the objective difference between something that could be explained by convergent evolution, horizontal gene transfer or “mistakes in the interpretation of the data” and something that would falsify common descent?

What is the objective difference between a mammal with feathers or any of the examples that I posted before?

¿Why can´t a feathered mammal be explained by convergent evolution, or by simply changing the phalogenetic tree (putting feathers before mammals and birds diverged) or by any other of the mechanisms that would be consistent with common descent?


For any creationist model to work, we'd need either a definition of what kinds are or at least a definite model on how to distinguish between kinds. In other words, how can you be sure that the relationship between organisms stops?



Well if there are significant discordances between 2 organisms then they can´t be related, for example I wouldn´t expect to see any discordance when comparing black humans and white humans (because they are the same kind), but I would expect discordances between humans and chimps/gorillas.
Wed Jul 02, 2014 10:04 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

Dandan wrote:Ok, I will make some statements and please tell me if you agree with them, this will help me to identify the exact points of disagreement.


Sure, although I think I've already answered them all.
Also, I'm a bit peeved that you won't answer the questions I put to you.
Nevertheless...

Dandan wrote:1) Some tree has to predominate regardless if evolution, creation or some other model is true.

With “tree that predominates” I mean that when comparing the genome in different organisms, some of them would necessary be more similar to others.

Agree? If not please tell me specifically why do you disagree.


Wrong.
I'll give you an answer for each of the three possibilities.

If evolution is correct, then I would expect one "perfect" tree, but I would also expect that humans can never find that one perfect tree due to the limitations already explained above. This would also only be true for organisms that share a common ancestor! If ever aliens were to visit earth, I'd expect them to share very little in common with us. In that case, we'd have one tree for earthlings and one for aliens. It would completely (and quite literally) blow my mind if they shared much the same underlying structure with us or if they could interbreed with us, a certain impossibility if ever there was one.

If some other model is true, then I don't know what would happen because I don't know what that model would be. If there is no common ancestor of all living organisms on earth, then maybe there are two or three. We needn't consider that idea because there would still be a tree of some kind at the root of every organisms history. There would still be one tree for group A, one for group B, etc.

If creationism is true, I would expect much the same. I would expect that there is one tree for every "kind", whatever that may be. However, I would also expect that multiple kinds show no relation to one another. I'll explain that in a second. What I certainly would expect though, is that multiple trees brought together would be at the same distance from one another and there are multiple trees, each with the exact likelihood as the other.

For example, let's take three groups of animals that creationists would consider "kinds" and show how they should match up if creationism were true. Creationist "papers" (I'll give just one reference) often list the following "kinds": Bovinae (Cattle kind), Sciuridae (Squirrel kind) and I'll add birds as another kind, though I don't know if they're considered one. It doesn't really matter, what really matters is their relationship. If I were to construct a phylogenetic tree of those three, I'd expect ((Bov, Sci)Bir) and nothing else.
But if creationism is true, they might as well be classified ((Bov,Bir)Sci) or ((Sci,Bir)Bov) or (Sci,Bir,Bov) or any other way. None of the trees would be any more likely than the other, because they wouldn't be related.

The opposite is true, of course. I showed the Lurasiatheria paper above, which clearly lists Rodents and Bovinae inside of Boreoeutheria, again inside Mammalia, which is a sister clade to Reptilia.

Now I said that "kinds" should show no relationship to one another. If life truly were designed, then I wouldn't expect organisms to share DNA and I certainly wouldn't expect a bird virus to adapt to a human host. That makes no sense. It's almost as if a designer had purposely built life so that it looks related. That's weird, isn't it?

Dandan wrote:2) If for example orangutans would have been more similar to humans than chimps, you would have drawn a different tree with humans and orangutans in the same clade, and you would have put chimps in a “distant clade” and you would have call that the correct tree. That wouldn´t falsify common descent, it would simply chance the order in which different organisms diverged.

Agree? If not please tell me specifically why do you disagree.


I already explained that above, don't you remember?

Inferno wrote:Life either evolved exactly one way or it didn't evolve at all. There is little point in asking "what if life evolved differently?" because it can't inform on real life. You're asking Sheldon Cooper's question "Have you ever wondered how humans would be different if they evolved from lizards instead of mammals?" Some people think about that stuff, but why?


If life had evolved differently then yes, the tree would be drawn differently. But it hasn't so the tree remains the same.

Dandan wrote:3) A concordant nested hierarchy, with statistically insignificant discordances would prove common descent with near 100% certainty, and ID and other alternative models would be wrong or latest inferior to common descent.

Agree? If not please tell me specifically why do you disagree.


Yes. Also, ID/creationism would be absolutely wrong.

Dandan wrote:4) A discordant nested hierarchy, with statistically significant discordances would falsify common descent and ID would be consistent with the data. In other words in a ID model any attempt to organize organisms in a Nested Hierarchy would be full of discordances.
Just to be clear A discordant tree is when a portion of DNA does not match the predominant tree, for example if the predominant tree puts humans more closer to chimps than to mice, a discordant gene would be a gene that puts humans closer to mice than to chimps.

Agree? If not please tell me specifically why do you disagree.


Yes to the first, absolute and resounding NO to the second.
As I already explained, the gene tree doesn't always match up with the species tree, that is to be expected. Incomplete lineage sorting, as explained above. Remember?

Inferno wrote:I will, however, take the time to link you to the following blog post over at BioLogos. I'm not particularly fond of them, but the post is excellent.
It explains why gene trees are not always exactly the same as the species trees and it's the species trees we're concerned with. (That directly fouls papers 1, 3, 4 and 5) This also explains why the rest of your argument is invalid: You're looking for discordant gene trees, but they're to be expected. That's why you look at multiple loci instead of just one and that's also why you look at multiple lines of evidence.

As Isotelus pointed out, "[d]iscordance would falsify evolution if studies showed both consistent and significant mismatches between gene and species trees". You're showing individual discordant genes, but even they fit into the overall lines established. That's what I showed when I provided the Laurasiatheria paper: Multiple lines of evidence with scientists looking at multiple genes provide the evidence, singular genes might not reflect that due to incomplete lineage sorting.


So yes, if the discordance were both significant and consistent, that would be a problem. Individual genes, on the other hand, won't phase me.

Dandan wrote:So if you if you agree with all these points, then the only point of disagreement is if the discordances that we observe are statistically significant.


I don't agree with all of those points, not by a long shot. But sure, let's agree that the only really important bit is the statistical significance of the discordances.

Dandan wrote:What objective criteria are you using to determine that a mammal with feathers would falsify common decent, but something like a platypuss with “snake venom” won´t falsify it? What is fundamentally different between these 2 examples?


I already explained both of those above. I specifically said that the article you linked to does not show that platypuses have snake venom, that's simply not true. Instead, the authors found evidence of convergent evolution. That's not the same. And I explained that above, look:
"Number three already has the answer in the line you quote: "resemble". This means "similar, but not the same". I didn't even have to open the nature article or read the study to know that they were talking about convergent evolution. Turns out I was right: "Poisonous platypuses confirm convergent evolution"
The paper mentions that many of the proteins are functional analogues, meaning that they are structurally different but have similar functions."

However, if the genes were exactly 1:1 the same, or if a significant portion of them were, then we'd have a serious problem.

This also gives me a chance to rectify an oversight I made. I don't read the peanut gallery so I have no idea what people are writing, but I forgot to mention that certain proto-feathers have emerged elsewhere, I believe pterosaurs had them. Those are again examples of convergent evolution, not of the exact same thing evolving again.
If avian feathers (and wings) suddenly sprouted on the back of a human, that would be the nail in the coffin of evolution. The below... would be impossible.

Image

So my answer is this: I would use any statistical p-value test (chi-squared or student's t-test, whichever suits the subject studied) and explore whether convergent evolution is statistically significant. (p<0.001) Anything with 0.1<p would confirm your hypothesis. I already explained that, too!

So to answer your question "what is the difference between the two examples" I would say the following: You don't understand the first example, but you're even more wrong about the second. That's the only difference.

Dandan wrote:Under what objective criteria do you say that 30% discordances in the human/Chimp/Gorilla genome is insignificant? What if instead of 30% we would have had 35% or 40% or 50% up to what point would you say that these discordances are significant?


I'm at a loss here. I already linked back to the explanations provided by Isotelus because she already gave answers in astonishing detail. What more do you want?

I'll try to give you an answer, but it's not easy to do without again resorting to the p-value answer I gave you above.

The following is a phylogenetic tree for monkeys.

Image

Website wrote:The picture shows a phylogenetic tree generated using noncoding DNA sequences (min. 2 kb per species; average of 6.5 kb per species)
in PAUP (both Maximum likelihood and maximum parsimony). Numbers above lines represent the Bootstrap percentages out of 2000 replications that that particular node was supported. Numbers below lines represent the number of synapomorphies that would need to be changed to alter the grouping.


In that picture and in that paragraph is basically everything you need. I've also contacted the author to see if I can further clarify the issue.

Dandan wrote:In other words, what is the objective difference between something that could be explained by convergent evolution, horizontal gene transfer or “mistakes in the interpretation of the data” and something that would falsify common descent?

What is the objective difference between a mammal with feathers or any of the examples that I posted before?


I already answered both of those questions in my earlier posts. If you can't be bothered to read my replies, then I can't be bothered to write them.

Dandan wrote:¿Why can´t a feathered mammal be explained by convergent evolution, or by simply changing the phalogenetic tree (putting feathers before mammals and birds diverged) or by any other of the mechanisms that would be consistent with common descent?


I've explained that too, but it's worth repeating.

"What would happen if mammals with feathers were found? As far as I know, but I might be wrong on that, feathers have only been found in theropod dinosaurs and their descendants: Birds. If their exact feathers, with the same structural development and genetic coding were found on one mammal, but not on any other mammal, then we'd have a serious problem. I'd go as far as to say: That would utterly disprove evolution."

AronRa explained this both in his Falsifying Phylogeny and in his Falsifying Phylogeny II: Evolutionary Law. Birds evolved feathers and they're nested among archosaurs and then reptiles and eventually chordates. They haven't shared a common ancestor with us for hundreds of millions of years. If we were to find two structures or traits evolving in very distantly related lineages, like humans suddenly sprouting wings and feathers, that would reduce evolution to horse feathers.
Note that this isn't the same as convergent evolution, where similar traits and structures are evolved, but they can be (easily) distinguished from one another due to their morphologic and/or genetic make-up. I explained elsewhere how we can determine that wolves and thylacines are the result of convergent evolution.

If, on the other hand, we evolved completely different feathers, maybe because the hair on our arms were to thicken, grow outward and straighten, that would be something completely different because the underlying structure isn't the same. I'd consider it unlikely, but not at all impossible.

Dandan wrote:Well if there are significant discordances between 2 organisms then they can´t be related, for example I wouldn´t expect to see any discordance when comparing black humans and white humans (because they are the same kind), but I would expect discordances between humans and chimps.


Then I'll fire that right back at you: How much? Give me a number and an equation, because apparently that's what's needed. Also, why wouldn't you expect what I expect, namely phylogenetic trees with the same statistical likelihood between multiple organisms?

----------------------------------------------------------------------------------------------------------

At this point, I would also like to state the following:
You don't seem to have read anything I wrote, otherwise you wouldn't be asking the same questions I answered in my first two posts. Writing the first two posts took a lot of time and effort, as you can see I'm already less inclined on my third post. If you can't be bothered to read what I write, then I won't be bothered to write what you don't read. If you repeatedly make me re-post what I already said, then I will put an early end to this debate. I simply won't have my time wasted in such a manner, though I am happy to provide if you show that you actually read what I write. Yes, I am that vain.
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Thu Jul 03, 2014 12:55 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

I don't agree with all of those points, not by a long shot. But sure, let's agree that the only really important bit is the statistical significance of the discordances.


Fantastic, thanks for making this long shot, so lets fosus on weather if the discordances that we see are significant or not.



I'm at a loss here. I already linked back to the explanations provided by Isotelus because she already gave answers in astonishing detail. What more do you want?

I'll try to give you an answer, but it's not easy to do without again resorting to the p-value answer I gave you above.

The following is a phylogenetic tree for monkeys.


You did not answer my question, isotelus didn’t answer the question in the post you linked, Aron didn’t answer it in any of the videos that you linked, and the nice image that you showed didn´t answered to the question ether.

I asked what % of discordance among the Chimp/Human/Gorilla genome would be sufficient to falsify common ancestry. We currently have a 30% discordance this means that:

70% of the genome shows this tree ((H,C)G)

15% shows this tree (H,G,)C)

And 15% shows this tree (C,G)H)

*For the sake of this particular point I over simplified what the authors really said

Since apparently this combintatoin of 70%,15%,15% is not a problem for you, then what combination would represent a problem? Would 80% 10% 10% represent a problem? Would 40% 30% 30% represent a problem? Up to what point would you say that the discordances are statistically significant?
I have the impression that it doesn´t matter what discordances we find, you will always explain them in a context of “common ancestry”


So my answer is this: I would use any statistical p-value test (chi-squared or student's t-test, whichever suits the subject studied) and explore whether convergent evolution is statistically significant. (p<0.001) Anything with 0.1<p would confirm your hypothesis. I already explained that, too!



Ok, so I what to see the numbers, I what to see why would a mammals with feathers would falsify common ancestry, I what your numbers to remove the possibility of convergent evolution, and your numbers also have to remove the possibility that feathers evolved before mammals and birds diverged.

I already explained both of those above. I specifically said that the article you linked to does not show that platypuses have snake venom, that's simply not true. Instead, the authors found evidence of convergent evolution. That's not the same. And I explained that above, look:


That is not true, platypuses and snakes have the same proteins (same genes) for venom, but that is irrelevant you will still invoke convergent evolution as you did with the sonar in whales and bats.

I what to see your numbers, I what to see the numbers that show that these examples can be explained by convergent evolution but that a mammals with feathers can´t.
AronRa explained this both in his Falsifying Phylogeny and in his Falsifying Phylogeny II: Evolutionary Law. Birds evolved feathers and they're nested among archosaurs and then reptiles and eventually chordates. They haven't shared a common ancestor with us for hundreds of millions of years. If we were to find two structures or traits evolving in very distantly related lineages, like humans suddenly sprouting wings and feathers, that would reduce evolution to horse feathers.



No he did not explain that, you are not answering my question;n I what to know what would prevent you from concluding that feathers evolved before mammals and birds diverged, if mammals some would have had feathers


Then I'll fire that right back at you: How much? Give me a number and an equation, because apparently that's what's needed. Also, why wouldn't you expect what I expect, namely phylogenetic trees with the same statistical likelihood between multiple organisms?



Sure I can provide the equations, but keep in mind the following, you are expected to provide your own equation (or model) and then explain why your equation is better than mine, if you refute my equation but you don´t substitute it for an equation that proves your point, we will end up with “agnosticism” or in other words we will end up with “We don´t know” (we don´t know if the discordances are statistically significant) and therefore we would have to conclude that we don´t know if common ancestry is the best model.

In other words, even if you refute my equation, you are still expected to provide an equation that proves that the discordances that we are not statistically significant, as oppose to something like a mammal with feathers.

Ok so since all human races belong to the same kind and chimps belong to a different kind, we wouldn´t expect to find a “chimp-trade” in Australian Aborigines, that other humans lack.

So for example if aborigines and chimps share a protein, or a portion of DNA or whatever with 15 bases pairs that is absent in the rest of the human population, that would imply that this protein “evolved” in the aborigine population AFTER they diverged from other human races.

So lets calculate the odds for such an even to take place, but before lets make some valid assumtions

Mutation rate = 100 mutations per generation

Genome size = 3,000,000,000 base pairs

Population size = 1,000,000 Australian aborígenes

1 human generation = 20 years

So with those assumptions, If the mutation rate is 100 and the genome isize in 3,000,000,000 base pairs, then the chances of having a mutation on a specific location is 1 in 30,000,000 (it has to be on a specific location, because the mutation has to be homologous to an existent mutation in the chimp genome)

Since the population size is 1,000,000,000 then if you divide 30,000,000 with 1,000,000 you have 30….1 in 30 are the chances of getting such mutation in at least 1 individual, or in other words you need 30 generations (600 years) to achieve such mutation.

The probability of achieving fixation in the aborigen population is 1 in 2,000,000,000

So 2,000,000,000 x 30 = 60,000,000 this is the number of generations that you need

So the chances of getting 1 homologous mutation in 2 unrelated clades is 1 in 60,000,000 in my example we are using a protein with 15 base pairs, so the actual number is 60,000,000 x 15 = 900,000,000 generations

Since 1 generation = 20 years we need 900,000, x 20 = 18,000,000, years, since we only have 6,000 years in my creationist model the chances of getting a protein in aborigines and chimps is 1 in 3,000,000. (18,000,000, / 6,000)

So in my model the probability of having a 20 bias pared protein in two unrelated animals, is 1 in 3,000,000. If you randomly look for discordances in a sample of 30,000 “kinds” you have a 1% chance to find a discordance. Discordances are not impossible but very unlikely.


Under your model all you have all you have to do is substitute 6,000 for the number of years that you consider accurate (for example 5,000,000 years) when comparing the chimp/human/gorilla genome

Under your view the probabilities of having 1 discordance among the human/gorilla/chimp genome is 18,000,000/5,000,000 or 27%
If evolution where true there would be a 27% chance of having 1 discordance in the chimp/human/gorilla genome or a 7% of having 2 discordances or a 2% chance of having 3 discordances, since there are thousands of discordances (30% of the genome) the chances for having so many discordances is virtually 0% according to my statistical method, but feel free to provide your statistical method.
Thu Jul 03, 2014 7:23 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

BTW

This is the actual ecuation


Mt Ps Pf Nb
-------------------
Gs

Mt = Mutatoin rate

Ps = population size

Pf = probability of achiving fixiation

Nb = number of base pairs that the homologous protein has

Gs = Genome size (in base pairs)


The result represents the probability of having to homologous proteins in unrelated clades by unguided mechanisms, if you divide the resoult by the number of generations (or tries) you have the actual probability of such an even occurring in a given amount of time.

So know that you know the equation, please prove to me that the discordances the we observe are statistically insignificant. If you final resoiult is less that 50% this means that common ancestry is more likely to be wrong than correct
Fri Jul 04, 2014 3:52 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

So once again, pretty much half of my post got ignored?

Dandan wrote:You did not answer my question, isotelus didn’t answer the question in the post you linked, Aron didn’t answer it in any of the videos that you linked, and the nice image that you showed didn´t answered to the question ether.

I asked what % of discordance among the Chimp/Human/Gorilla genome would be sufficient to falsify common ancestry. We currently have a 30% discordance this means that:

70% of the genome shows this tree ((H,C)G)

15% shows this tree (H,G,)C)

And 15% shows this tree (C,G)H)

*For the sake of this particular point I over simplified what the authors really said

Since apparently this combintatoin of 70%,15%,15% is not a problem for you, then what combination would represent a problem? Would 80% 10% 10% represent a problem? Would 40% 30% 30% represent a problem? Up to what point would you say that the discordances are statistically significant?
I have the impression that it doesn´t matter what discordances we find, you will always explain them in a context of “common ancestry”


First, at no point do the authors say anything about 70/15/15, that's simply wrong.

Second, I already told you that I know very little about statistics so I'd be hard-pressed to find you a number. I can only repeat what the scientists found. I am taking a course in advanced biological statistics, so feel free to drop me a line in 2-3 months from now if you really want an answer.

Now, as for the scientists answer: They predicted the values using a software (CoalHMM) and then crunched the numbers. The numbers matched up with the prediction. As Isotelus already stated, they predicted 0.9% of orangutan alleles would be more similar to humans and the actual number they found was 0.8%. Now tell me, how can something that's predicted by evolution also falsify evolution? You're making no sense.

Now I wouldn't expect the numbers to deviate too much from the expected value, but I don't know how to calculate it. All I can do is once again refer you to the probability I alluded to above: Anything with a low p-value (p<0.001) would support "my" hypothesis, anything with p>0.1 would confirm your hypothesis. Well no, actually that would just mean that "my" hypothesis isn't correct.

In any case, how can I reasonably calculate what you're asking me for? I'd have to know my way around CoalHMM and know a large deal of statistics. You're in no way being reasonable in your expectations. Know that it's taking trained scientists months to figure this stuff out and you're expecting me to do it in an afternoon?
But I already proved that I don't need to do that because we're already seeing a great deal of concordance between species that, according to you, are supposed to be unrelated. If your pet hypothesis were true, we wouldn't expect to find that ever.

Dandan wrote:Ok, so I what to see the numbers, I what to see why would a mammals with feathers would falsify common ancestry, I what your numbers to remove the possibility of convergent evolution, and your numbers also have to remove the possibility that feathers evolved before mammals and birds diverged.


You can refer to the numbers I provided below. To see an exact replica of a feather would probably take more than 15 BPs, so the odds of two separate populations coming up with the exact same mutations are astronomical. My calculator gave me an error message when I tried to compute it, so I'll make a vague guess: If you had living things for the duration of the whole universe, or roughly 13.5 billion years, and you had two unrelated lineages which are both identical in their original make up (meaning organic, DNA/RNA protein-based, etc.) and both have similar living conditions, and they started to diversify, then after these 13.5 billion years you still wouldn't have them both produce the same animal independently. That's roughly how improbable it is for humans to grow bird feathers. (Without genetic engineering, of course)

Dandan wrote:That is not true, platypuses and snakes have the same proteins (same genes) for venom, but that is irrelevant you will still invoke convergent evolution as you did with the sonar in whales and bats.


Wrong.

Novel venom gene discovery in the platypus wrote:It is now clear that the venom of the platypus contains a diverse range of proteins, many of which may be functional analogues of venom components of other species, including reptiles, insectivores, fish, and even invertebrates. Reptiles diverged from the vertebrate lineage 315 million years ago, and platypuses diverged from the rest of the mammals 166 million years ago [5]. The fact that these extremely divergent species share similar venom components, some of which were found repeatedly in platypus and other venoms, suggests that there are indeed protein motifs that are preferentially selected for independent evolution to venom molecules in a striking display of convergent evolution, and that many animal venoms share some similarities in their mode of action [27].


I even explained that:
Inferno wrote:Number three already has the answer in the line you quote: "resemble". This means "similar, but not the same". I didn't even have to open the nature article or read the study to know that they were talking about convergent evolution. Turns out I was right: "Poisonous platypuses confirm convergent evolution"
The paper mentions that many of the proteins are functional analogues, meaning that they are structurally different but have similar functions.

The paper goes on to note that the venoms evolved independently:
"The fact that these extremely divergent species share similar venom components, some of which were found repeatedly in platypus and other venoms, suggests that there are indeed protein motifs that are preferentially selected for independent evolution to venom molecules in a striking display of convergent evolution, and that many animal venoms share some similarities in their mode of action [27]."


If you can't accept the evidence, then there's no need to blame me for your failures. I'm not invoking anything, I'm simply repeating what scientists have found. Deal with it.

Dandan wrote:I what to see your numbers, I what to see the numbers that show that these examples can be explained by convergent evolution but that a mammals with feathers can´t.


I already explained that your premise is wrong, so your conclusion makes no sense. If the same genes for snake venom were also found in monotremes, you might be on to something. However, the genes are NOT the same, they're functional analogues. They produce similar proteins, but they're not the same genes. Provide a quote where it actually says that they're the same.

Dandan wrote:Equations and stuff


First, I'm truly baffled by your equations. I don't know where you got any of them from, but I'm pretty sure they're complete hogwash. In fact, I checked several textbooks to make sure and I was right: Not a single textbook uses the equations you use. I'll use Zackay (2007). (This seems to be seminar literature in Germany and it is concordant with all other literature I browsed.)
There are three equations we need:
First, the effective population size.
Image

Second, the fixation time for a mutation with no selective advantage.
Image

And third, the probability for fixation:
Image

Nf = number of females, Nm = number of males, Ne = effective population (I'll simplify vastly and use Nf and Nm as 500,000 each)
G = Generation time (20 years is reasonable)
N = population size

So for Ne = (4 x 500,000 x 500,000)/(500,000+500,000) we get Ne = 1,000,000 or 10^6

For the fixation time, we have t = 4 x 10^6 x 20 so t = 80,000,000 or 80 million years.

We don't need to find out the probability for fixation, but I'll give it to you: 0.0000005 or 0.00005%, also known as 1:2,000,000

What I've calculated here is the probability and time it would take for one specific mutation to spread through and become fixed in a population. The time for the mutation to occur is rather quick:
Mutation rate x number of sites to mutate x number of copies of the gene, which is
100/3,000,000,000 x 15 x 1 when using your example or 1:2,000,000 chance. With a growth rate of 1% (very reasonable) we'd have 10,000 kids so 10,000x1/2,000,000 = 0.005 or 5 mutations every 1000 years. That's neglegible next to our 80 million years.

Note that the time for fixation with a moderate selection coefficient of s=0.01 is a mere 58,000 years, less so if we assume a generation time of 15 years, which is vastly more reasonable for a population a few thousand years ago. (Which is what I assume we're talking about?)

So already we can see that the equations you provide don't have anything to do with what you're talking about. I'm especially freaked out by the way you calculate probabilities.




Second, your numbers vary from line to line. One time, the population size is 10^6, one time it's 10^9. You're confusing offspring with population size (hint: Populations don't double every year, not human populations anyway), you incorrectly state that the probability of achieving fixation is 1:2,000,000,000 when it's actually 1:2,000,000. One time you compute 900,000,000 generations, seconds later it's only 900,000 generations.

I also have no idea how you calculate the probability of discordance, it makes absolutely no sense.



In any case, this is all absurd to the highest degree. What you actually calculated (and where I provided the correct calculations) is the probability and time for a mutation to arise in independent populations. That is to say, I would be really surprised to see two identical mutations in two non-related species. If your creation-model is correct, I'd expect two unrelated organisms (chimp and human, for example) to have very few mutations in common. Very few indeed.

Instead, we find this:
Image

This is a picture of a gene (aquaporin 10) comparison between Homo Sapiens and Pan Troglodytes using BLAST. I located a random portion of human DNA (on chromosome number 1) and searched for identical or nearly identical portions in chimp (Pan Troglodyte) DNA. This is only one of several thousand possible gene matches I got, and this is only a portion of that gene, but it will suffice.

As you can see, 1043 out of 1062 BPs match, the odds of which I don't know how to calculate but are sure to be astonishing. (I came up with 1:several billion, but I'm fairly sure my math is off...) How about you deal with that?
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Sun Jul 06, 2014 12:26 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

So once again, pretty much half of my post got ignored?


I think we´ve made very important progresses in this debate. At this point we both agree that we do find discordances, being our main point of disagreement weather if this discordances are statistically significant or not.

We also agree that it would be very unlikely (nearly impossible) for two animals to share genes or portions of DNA if they are not related, since it would be very unlikely for two animals to suffer from the exact same mutations.

With that said, have I ignored anything relevant? I know that I have ignored some stuff, but that is because I ether agree with you or because I considered it irrelevant.


First, at no point do the authors say anything about 70/15/15, that's simply wrong.

Second, I already told you that I know very little about statistics so I'd be hard-pressed to find you a number. I can only repeat what the scientists found. I am taking a course in advanced biological statistics, so feel free to drop me a line in 2-3 months from now if you really want an answer.

Now, as for the scientists answer: They predicted the values using a software (CoalHMM) and then crunched the numbers. The numbers matched up with the prediction. As Isotelus already stated, they predicted 0.9% of orangutan alleles would be more similar to humans and the actual number they found was 0.8%. Now tell me, how can something that's predicted by evolution also falsify evolution? You're making no sense.

Now I wouldn't expect the numbers to deviate too much from the expected value, but I don't know how to calculate it. All I can do is once again refer you to the probability I alluded to above: Anything with a low p-value (p<0.001) would support "my" hypothesis, anything with p>0.1 would confirm your hypothesis. Well no, actually that would just mean that "my" hypothesis isn't correct.

In any case, how can I reasonably calculate what you're asking me for? I'd have to know my way around CoalHMM and know a large deal of statistics. You're in no way being reasonable in your expectations. Know that it's taking trained scientists months to figure this stuff out and you're expecting me to do it in an afternoon?
But I already proved that I don't need to do that because we're already seeing a great deal of concordance between species that, according to you, are supposed to be unrelated. If your pet hypothesis were true, we wouldn't expect to find that ever.



Ok what the article really says is that the ((H,C)G) tree is predominant 70% of the time, and 30% of the time some other tree predominates. My question is if a discordance of 30% is not enough to falsify common descent, what % of discordance would falsify it?

quote form the article
quote]In 30% of the genome, gorilla is closer to human or chimpanzee than the latter are to each other
[/quote]


I honestly don´t think that the question was answered, you mentioned an Orangutan paper, and I honestly don´t understand why is that relevant if I am talking about the discordances in humans chimps and gorillas. As for the orangutan paper, at most you are showing that in this particular case we do see a concordant tree, I am not denying that sometimes we do find concordances.

You can refer to the numbers I provided below. To see an exact replica of a feather would probably take more than 15 BPs, so the odds of two separate populations coming up with the exact same mutations are astronomical


Well that is a remarkable claim, lets look at the echolocation in bats and dolphins.


Using maximum likelihood (ML),
maximum parsimony (MP), and
neighbor-joining (NJ) methods [5], we
reconstructed the prestin protein tree.
Surprisingly, all methods group the
dolphin within the bats — specifically
with the green-labeled families rather than with the cow,
its true closest relative in our dat

we here report the
surprising finding that the bottlenose
dolphin, a toothed whale, is clustered
with microbats in the gene tree

http://ac.els-cdn.com/S0960982209020570 ... aa5c8b3a24

Paraphrasing what the papers says; what the researchers found was that when you do a family tree with the gene pristine dolphins appear more closely related to bats than to whales. In your comment you are implying that it would be unlikely for a mammal to evolve a feather (15 BPs) because the chances would be very small, well in this case we have 9 base pairs I am sure that if you do your calculations, you will also get an astronomically small number.

The authors explain this by arguing that bats and dolphins received the exact same mutations independently, exactly what you said was impossible.

So even though I didn’t present a mammal with feathers I presented something equivalent


About the platypus
The paper goes on to note that the venoms evolved independently:
"The fact that these extremely divergent species share similar venom components, some of which were found repeatedly in platypus and other venoms, suggests that there are indeed protein motifs that are preferentially selected for independent evolution to venom molecules in a striking display of convergent evolution, and that many animal venoms share some similarities in their mode of action [27



I think you are misunderstanding what the authors are saying, when they talk about convergent evolution, they mean that the same genes evolved independently,( in other words convergent evolution at a genetic level,) this is not analogous to the alleged convergence between wings in birds and bats, where similarities are only superficial. In the case of the venom in platypuses and snakes we are talking about the same genes. Supposedly the genes evolved twice independently, like in the case of pristine in bats and dolphins the same genetic material evolved independently more than once, exactly what you said was impossible.

quote about the platypus and its venome:

We identified 83 novel putative platypus venom genes from 13 toxin families, which are homologous to known toxins from a wide range of vertebrates (fish, reptiles, insectivores) and invertebrates (spiders, sea anemones, starfish).


Both the bat and the platypus examples are examples of organisms evolving the same genetic material independently; exactly what you said was impossible. And just to be clear these are not isolated examples, there are thousands of similar cases.

About the ecuations:


The whole point I was trying to make is to show that it would be very unlikely for two unrelated clades to suffer from the exact same mutations, since we both agree with that conclusion I see no point in discussing on weather if my equations are correct or not

In a private message I asked you this:
Inferno, In my equation I calculated the probability of 2 genes* evolving independently in 2 distantly related clades, and concluded that the chances are virtually cero. Regardless if you accept my numbers of not, do you at least accept the conclusion? Do we agree on the fact that it is extremely unlikely to evolve the same genes independently in 2 unrelated clades?

And you answered this.
Yes, I absolutely agree with that conclusion. As I said in the post:

"What you actually calculated (and where I provided the correct calculations) is the probability and time for a mutation to arise in independent populations. That is to say, I would be really surprised to see two identical mutations in two non-related species."

The only thing I did was show that both your numbers and equations were in error and that the more-or-less correct calculations are the ones I provided.

What you did NOT provide was any measure of "discordance" or anything like that.


The whole point I was trying to make is to show that it would be very unlikely for two unrelated clades to suffer from the exact same mutations, since we both agree with that conclusion I see no point in discussing on weather if my equations are correct or not

So basically you said that two unrelated clades can´t have shared mutations, and I did showed you examples of shared mutations among clades that are supposed to be unrelated. ¿can I claim vicory?

As for your equations I have nothing to say about them, I accept them as accurate.

In any case, this is all absurd to the highest degree. What you actually calculated (and where I provided the correct calculations) is the probability and time for a mutation to arise in independent populations. That is to say, I would be really surprised to see two identical mutations in two non-related species. If your creation-model is correct, I'd expect two unrelated organisms (chimp and human, for example) to have very few mutations in common. Very few indeed


In fact I do agree with you, but I don´t think humans and chimps have any shared mutations, they have traits in common, but this are not due to shared mutations, nor due to common ancestry, but rather by “common designer”

You see chimps and humans have similar needs (digest food, breathe, pump blood etc.) it makes sense to say that a designer would design similar systems to satisfy similar needs, and It would also make sense to say that a designer would create similar genes to design similar systems.

For example Microsoft Word and Microsoft Power Point have a similar spelling check system, and I am sure that this system is encoded by similar codes in both programs, after all if both systems need a spelling check system, why wouldn’t the designer use the same code in both of the systems?

Same thing with DNA, if humans and chimps need to breathe the same air, why wouldn’t a designer create them with a similar respiratory system with similar genes?

In my model similarities among chimps and humans are not due to shared mutations,
Mon Jul 07, 2014 7:16 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

I absolutely refuse.

I posted an in-depth explanation as to why your "equations" are wrong and you didn't even acknowledge it. I gave you a sequence of DNA shared by Chimps and Humans that, according to you, we should not find. You didn't acknowledge that. I explained that the genes found in platypus and snake venom are not the same, instead they share similarities and code for similar genes, aka they are functional analogues. Yet you still repeated the same mistake. I'll come back to that in a later post, once you've answered the following questions. I regard all of these important to some degree, as you have either ignored half of my posts (while I responded to your every word!) or because we seem to be of vastly different opinions on their answers.

With that said, before I respond to your post, I have to press you on the many (and in my opinion important) points I raised before:
      Do you agree that there are potentially billions of trees, but that we hit upon a very limited number of trees (within margins of error I discussed above), something that is very unlikely?
      Do you further agree that, if creationism were true we shouldn't find this at all, that instead we should find many trees, all with the same likelihood?
      Do you also agree that the branching pattern found can only be explained by evolution, not by anything else?
      Do you agree that not a single test has shown obvious problems (like dogs being more closely related to ants than to humans) even though that's what we'd expect if creationism were true? If you don't, provide the study where that is the case.
      Do you agree that blind tests of phylogenetic analysis have shown the algorithms to hit upon the right tree in every case?
      Do you further agree that the methodology and algorithms behind phylogenetic analysis are rigid and reliable?
      Do you agree that organisms creationists would call "created kinds" or "baramins" are, contrary to their claims, definitely related? (This is what the cladogram and Zhou et al. (2012) was all about...)
      As an aside, do you agree that even the methodology presented by creationists falsifies their own ideas? Meaning that creationist methodology also definitely shows that organisms they think are unrelated... are really related to everything else.
      Do you agree that I gave a sufficient answer as to why avian feathers on humans would absolutely falsify evolution?
      Do you agree that all bootstrap values and p-values we have so far looked at show an extremely high level of confidence in favour of evolution?
      Do you agree that I gave an in-depth and comprehensive explanation as to why your equations are wrong and that mine are at least approaching accuracy?
      Do you agree that the amount of matching DNA between chimps and humans is both predicted by evolution and predicted in the real world?
      Do you further agree that the only explanation of this fact by creationists is an ad-hoc rationalisation, aka. "common design, common designer"? *

*Note: I'll add an explanation to that in my next post. For now, I want my questions answered as I believe the amount of work put into the posts justify them.
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Tue Jul 08, 2014 6:41 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

.

With that said, before I respond to your post, I have to press you on the many (and in my opinion important) points I raised before:
      Do you agree that there are potentially billions of trees, but that we hit upon a very limited number of trees (within margins of error I discussed above), something that is very unlikely?




Yes there are billions of possible trees, however of all the trees one has to be the predominant tree regardless if creation, evolution or some other model where true, in this case the tree that predominates is the (Human, chimp)gorilla) Orangutan) Gibbon)Monkey) etc.) in some other parallel universe another tree would be the predominant tree.

My point is that if you analyze the whole genome some tree has to predominate, this is necessary true regardless of the model. However each individual gene has a different predominant tree, for example pristine puts dolphins more closely related to bats than to whales.

If evoluton where true, we would expect to have a predominant tree and each individual gene should follow that pattern.
If creation where true, we would expect to find a predominant tree, but each gene would have a different tree




    Do you further agree that, if creationism were true we shouldn't find this at all, that instead we should find many trees, all with the same likelihood?[/
list]

That is not necessarily true, a designer could have created Kind A more similar to Kind B than to Kind C.
For example a computer engineer can create computer A and Computer B similar to each other, and very different form computer C.



    Do you also agree that the branching pattern found can only be explained by evolution, not by anything else?


Yes a branching pattern could only be explained by evolution. It would be unparsimonious to say that God created life in a branching pattern as if evolution where true. However we don’t´see such pattern, that is my point, for example we see bats with mice-like hair and with dolphin-like sonar


    Do you agree that not a single test has shown obvious problems (like dogs being more closely related to ants than to humans) even though that's what we'd expect if creationism were true? If you don't, provide the study where that is the case.


If dogs would have been “more related” to ants than to humans, then dogs would have had “ant genes”, dogs would have been ants. In other words dogs would have been classified in the ant clade.

However evolution predicts that ants and dogs should not have any traits in common unless this trait is also present in all vertebrates and in all the clades that descended from the common ancestor of dogs and ants; while creation would predict at least some exceptions, we will have to wait untilled more genomes are sequence to see who has the correct prediction. However I think we do have equivalent examples, like pristin being more similar in bats and dolphins than in dolphins and whales


    Do you agree that blind tests of phylogenetic analysis have shown the algorithms to hit upon the right tree in every case?


No, I don´t agree prestine for example puts bats and dolphins as close related and whales and dolphins as distant relatives contradicting the predominant tree. The same is true with most of the genes.

    Do you further agree that the methodology and algorithms behind phylogenetic analysis are rigid and reliable?


I don´t have problems with the test, my problem is the fact that phylogenetics presupposes common ancestry

    Do you agree that organisms creationists would call "created kinds" or "baramins" are, contrary to their claims, definitely related? (This is what the cladogram and Zhou et al. (2012) was all about...)


I did not undestrand the question

    As an aside, do you agree that even the methodology presented by creationists falsifies their own ideas? Meaning that creationist methodology also definitely shows that organisms they think are unrelated... are really related to everything else.


can you name an example



sometimes, you only presented 1 example (the orangután)
    Do you agree that I gave a sufficient answer as to why avian feathers on humans would absolutely falsify evolution?[/list
]

yes I agree, after

    Do you agree that all bootstrap values and p-values we have so far looked at show an extremely high level of confidence in favour of evolution?


I don´t think you ever provided a paper with P values, that actually explains at what point would a discordance would falsify evolution

    Do you agree that I gave an in-depth and comprehensive explanation as to why your equations are wrong and that mine are at least approaching accuracy?


I accidentally removed 3 ceros in the text, that was just a typing error, in my calculator I did used the correct values, so my result is still valid. That was my only mistake.
I simply calculated the chances for a 20 biased paird gene evolving independently in 2 different clades. Since we both agree that the probabilities are very small, I see no point in discussing the details.
I don´t disagree with you equations, I simply don’t ‘understand your point, as far as I understood you proved that mutations can become fixed once in a while, I don´t disagree.


    Do you agree that the amount of matching DNA between chimps and humans is both predicted by evolution and predicted in the real world?
    Do you further agree that the only explanation of this fact by creationists is an ad-hoc rationalisation, aka. "common design, common designer"? *



No, I don´t think that the 98% similarity (or whatever the number is) was predicted by evolution, I challenge to provide a paper that actually makes that prediction(remember predictions comes before the discovery)

Why wouldn’t the creator create 2 independent creatures with some traits in common? If this is common amoung human designers, why wouldn’t God or ET do the same? I don´t think this is an ad hoc


Please let me know if you agree with these 3 statements
1)A human with feathers would falsify evolution, because it would be very unlikely for humans to suffer from the exact same mutations that birds did
2)Something equivalent would also falsify evoluton
3)I did presented something equivalent, pristine in dolphins is more similar to bats than to whales, which would imply that dolphins suffered from the exact same mutations than bats independently





*Note: I'll add an explanation to that in my next post. For now, I want my questions answered as I believe the amount of work put into the posts justify them.[/quote]
Tue Jul 08, 2014 8:40 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

How annoying. I had a huge post written, then my computer installed updates. Forgot to save it of course. Sigh.

Dandan wrote:Yes there are billions of possible trees, however of all the trees one has to be the predominant tree regardless if creation, evolution or some other model where true, in this case the tree that predominates is the (Human, chimp)gorilla) Orangutan) Gibbon)Monkey) etc.) in some other parallel universe another tree would be the predominant tree.

My point is that if you analyze the whole genome some tree has to predominate, this is necessary true regardless of the model. However each individual gene has a different predominant tree, for example pristine puts dolphins more closely related to bats than to whales.

If evoluton where true, we would expect to have a predominant tree and each individual gene should follow that pattern.
If creation where true, we would expect to find a predominant tree, but each gene would have a different tree


Absolutely wrong. As I explained before, we'd expect one tree (with deviations) from evolution, but we'd expect multiple trees if creationism were true.

In his 2010 article "Uprooting the tree of life", Doolittle showed what was (and mostly still is) the consensus view. A "tree of life" should look something like this:
Image

Doolittle disagreed, claiming that it should look like the following:
Image

Some scientists disagree with Doolittle, noting that HGT probably doesn't play such a huge role, but that's minor quibbles. The two notions are certainly closely related and the truth probably lies somewhere in between.

Creation.com notes a third option: The so-called "orchard of life". As you can see, all trees are at a distance to another (many trees) and we can expect them to be of equal probability.
Image

That's why I tried to nail you on that, but apparently you don't even know stuff about your own beliefs.

Dandan wrote:That is not necessarily true, a designer could have created Kind A more similar to Kind B than to Kind C.
For example a computer engineer can create computer A and Computer B similar to each other, and very different form computer C.


This is basically the "same design, same designer" or "same genes, same designer" ploy. I'll deal with that below the fold.

Dandan wrote:If dogs would have been “more related” to ants than to humans, then dogs would have had “ant genes”, dogs would have been ants. In other words dogs would have been classified in the ant clade.


You're confusing the issue and I already dealt with that in my Simpsons-analogy.
Comparative anatomy predicted that dogs would be closer to humans than to ants, correct? This goes for millions of other species, but we'll keep it simple.
Now if creationism were correct and evolution were false, we might conclude that dog-genes are more similar to ant genes or that they're all at an equal distance. Refer to the orchard-hypothesis here. We don't because creationism is wrong.

Dandan wrote:No, I don´t agree prestine for example puts bats and dolphins as close related and whales and dolphins as distant relatives contradicting the predominant tree. The same is true with most of the genes.


That's not what I asked for. I specifically talked about the blind tests I mentioned twice before.

Inferno wrote:Now as an in-between, we need to know if cladistic analysis actually works. Will these analyses find the correct tree? As previously stated, they are already allowed in a court of law, but that's not good enough for me. What if we took some organism with a short generation span and tracked their lineages, that is to say we know exactly who the offspring of whom is. Out of over 130,000 possible trees, will cladistic analysis find the correct tree? As it turns out, yes it will.
TalkOrigins lists a few of these examples, so we know that cladistic analysis is damn accurate.


Do you agree that blind tests of phylogenetic analysis show that it is accurate? Do you agree that the probability of them finding that one correct tree is tantamount to zero? Do you agree that phylogenetic analysis also works in the real world?

Dandan wrote:I don´t have problems with the test, my problem is the fact that phylogenetics presupposes common ancestry


It was formally tested by Theobald in 2010, so why the fuss?

Dandan wrote:
Inferno wrote:Do you agree that organisms creationists would call "created kinds" or "baramins" are, contrary to their claims, definitely related? (This is what the cladogram and Zhou et al. (2012) was all about...)
I did not undestrand the question


Creationists typically assume that there are some "created kinds" or "baramins" like "dog kind" or "cat kind" or "ant kind" or "snake kind" or similar constructed groups. My point in showing the cladogram and the paper was to prove that, despite what creationists claim, these "kinds" are definitely related... and creationism is 100% wrong. As I pointed out, their relationships were found with maximum support.

Dandan wrote:
Inferno wrote:As an aside, do you agree that even the methodology presented by creationists falsifies their own ideas? Meaning that creationist methodology also definitely shows that organisms they think are unrelated... are really related to everything else.

can you name an example


I already did, you simply didn't read it.
Inferno wrote:Creationists are typically not taken seriously by scientists, but one scientist did: In 2010, P. Senter published an article in the Journal of Evolutionary Biology. In it, he used creationist methods to determine how (un)-related animals were. He found, unsurprisingly to scientists, that groups classified by creationists as distinct "kinds" were related even according to the models used by creationists.


Dandan wrote:
Inferno wrote:Do you agree that Incomplete Lineage Sorting (ILS) is both predicted and the predicted and observed match with an astonishing degree of accuracy?

sometimes, you only presented 1 example (the orangután)


That's not how it works. Either we can make accurate predictions, or we can't. An accurate prediction in the Orangutan genome was made, so you either agree that ILS is both predicted and observed or you ignore the facts.

Dandan wrote:I don´t think you ever provided a paper with P values, that actually explains at what point would a discordance would falsify evolution


Because that's not how it works and I explained that to you. I'd have to know a heck of a lot more in both statistics and biology, specifically genetics, to answer your question.
What both you and I can do, however, is look at the p- and bootstrap values. I explained in great detail that a p-value of p<0.001 is strong support for a hypothesis, while p>0.1 is strong support for the null hypothesis or rather against the hypothesis. Why do we constantly see p-values in support of common ancestry and evolution? You'd have to assume a huge conspiracy theory in science to weave that away.

Dandan wrote:I accidentally removed 3 ceros in the text, that was just a typing error, in my calculator I did used the correct values, so my result is still valid. That was my only mistake.
I simply calculated the chances for a 20 biased paird gene evolving independently in 2 different clades. Since we both agree that the probabilities are very small, I see no point in discussing the details.
I don´t disagree with you equations, I simply don’t ‘understand your point, as far as I understood you proved that mutations can become fixed once in a while, I don´t disagree.


Contrary to what creationists believe, accuracy matters. I'm not vexed about the three zeroes you added or removed, I saw that your calculations were more or less correct. What they are NOT is valid.
Your equations calculate... something, but they certainly don't calculate a probability.

Remember, I gave you the equation for the probability of fixation:
Image

It's 1:2,000,000

What you calculated is the time it needs for a mutation to arise and become fixed, only your equations are wrong.
If you want to calculate the area of a rectangle and you use the equation 2a+2b, everybody will laugh at you. That's the area for the perimeter, not the area. The equation for the area is a x b = A
I used your numbers but used the correct equation, that's why it's important to discuss the details.

I also didn't calculate the probability for fixation, I calculated the time it takes for a neutral mutation to arise (1,000 years) and to become fixed (80 million years). We both came up with a long time, yes, but as I said "accuracy matters".

What you don't understand is that it's the differences that matter, and I'll explain that in a second.

Dandan wrote:No, I don´t think that the 98% similarity (or whatever the number is) was predicted by evolution, I challenge to provide a paper that actually makes that prediction(remember predictions comes before the discovery)


I didn't say anything about "98% similarity" now, did I? I should have phrased that better and will explain that now. There are three examples I want to highlight:
Other great apes have 48 chromosomes, while we have 46. If we had simply lost a chromosome during our evolution, we would probably not exist. So there absolutely must be evidence of chromosomal fusion, otherwise we can’t explain how this miracle happened. Lo and behold, Human Chromosome 2 shows evidence of fusion. If a designer had designed the chromosomes, why didn't he do the same thing as with the other great apes? That’s what I would expect of a designer... but this? It can only be explained by evolution... or by a dumb designer.

Back in 1870, Darwin predicted that we should look for our common human ancestor in Africa. The Out of Africa theory (or "recent single-origin hypothesis", if we’re to be correct) is now more or less universally accepted.

Third, “almost all animals make Vitamin C inside their bodies. It was predicted that humans are descended from creatures that could do this, and that we had lost this ability. (There was a loss-of-function mutation, which didn't matter because our high-fruit diet was rich in Vitamin C.) When human DNA was studied, scientists found a gene which is just like the Vitamin C gene in dogs and cats. However, our copy has been turned off.”
This video explains it rather well.

These are just three predictions regarding the similarities between chimps and humans. On the site I linked to, there are a myriad other predictions, all of which turned out to be true. How can we make such accurate predictions if evolution is wrong?

----------------------------------------------------------------------------------------------------------

Dandan wrote:Ok what the article really says is that the ((H,C)G) tree is predominant 70% of the time, and 30% of the time some other tree predominates. My question is if a discordance of 30% is not enough to falsify common descent, what % of discordance would falsify it?


Didn't I explain, in the very portion you quoted, that I don't know? That I'd need to know a lot more than I know now, both about biology, specifically genetics, and statistics? I also explained that we, as laypeople, can check the p-values and look for incongruence there, but anything other than that is beyond my (and I assume your) ability.

Dandan wrote:I honestly don´t think that the question was answered, you mentioned an Orangutan paper, and I honestly don´t understand why is that relevant if I am talking about the discordances in humans chimps and gorillas. As for the orangutan paper, at most you are showing that in this particular case we do see a concordant tree, I am not denying that sometimes we do find concordances.


You don't think so, but I know that it was answered. As I pointed out, Isotelus gave you multiple, in-depth answers and I don't feel the need to revisit them. The information is right there, if you can't be bothered to read it then why should I write another page on it? My answers are already long enough, don't you think?

Isotelus wrote:The paper is dealing with discordant species trees only (the list of animal groups from the quotes similarly involve closely related species, most of which are in the same genera), which has no effect on the validity of congruence in higher-level phylogenetic ranks. Even if it did, the paper would still not be sufficient evidence for major discordance, as it is describing a statistical means by which to both predict and detect its incidence in gene and species trees: the key here is realizing that the mechanisms of discordance are well-understood. The paper is in effect stating that it is not actually a hindrance and can instead be used to determine the causes of its own occurrence, as well as extract information on ancestral populations. The paper describes the results found in some of the articles quoted above, and in spite of discordance, the gene trees could be cross-examined with other statistical methods such that they were able to resolve the relationships and divergence times between the species, as well as determine ancestral population sizes. Incidentally, one study included found that the (human+chimp)gorilla) tree was the most well-supported, which is a conclusion shown by other research and is the current majority consensus.


Dandan wrote:Well that is a remarkable claim, lets look at the echolocation in bats and dolphins.


Before I even read the paper, just from the portion you quoted, I already new the answer. You yourself quote that bats and whales were only clustered together in the gene tree for that particular gene (gene coding for a motor protein, as I understand it). Doesn't that remind us of an article I linked to before? I'm sure there was something...

Let's continue quoting the other, also relevant portions of the paper:
Indeed, the same misplacement of dolphin is observed in the Prestin tree reconstructed with only nonsynonymous nucleotide substitutions (Figure S1B); but, when only synonymous substitutions are used, dolphin and cow are correctly grouped with 100% bootstrap support (Figure S1C).


Now as I found out while searching, this gene "was present in the common ancestor of eukaryotes". If you look at the genes in cows and dogs, they don't appear to be all that different. If Zhou et al. (2012) (and later authors) were right, Fereuungulata are nested among Scrotifera (~63-66mya), and we know that Ceteceans took to the sea about 52mya. In that time, I'd not expect a huge deal of genetic difference, right? So we already have a very similar gene, that then diversifies slightly and then two animals have very similar needs: Finding where they are in a medium fit for sound waves and little light. That both hit upon echolocation, given the genes were already waiting, is hardly remarkable.

Convergent sequence evolution between echolocating bats and dolphins wrote:The ability of some bats and all toothed whales to produce sonar pulses and process the returning echoes for prey detection and orientation (echolocation) is a spectacular example of phenotypic convergence in mammals.


It further seems that the convergence is only regarding the phenotype, not the genotype. Now that would have been more remarkable. As I said before, prestin is a protein, so while we might expect the protein to remain the same, the amino acids have changed and the genes most certainly have.

Just for fun, here's the protein sequence for the genes in Humpback Whales and Leschenault's Rousettes (fruit bat). (The only relevant ones I could find on GenBank without going to too much trouble...) Compare their protein sequences. Yes, they're similar, but they're not the same.

The authors of the second paper even stated that outright:
Trees based on nucleotide alignments from this larger dataset strongly supported the accepted species tree topology, albeit with the clustering of echolocating bats reported earlier [3].


Just for those interested, here are the six relevant papers I could find. All except one Open-access.
Convergent sequence evolution between echolocating bats and dolphins (2010)
Cetaceans on a Molecular Fast Track to Ultrasonic Hearing (2010)
Molecular Evolution: Gene Convergence in Echolocating Mammals (2010)
The hearing gene Prestin unites echolocating bats and whales (2010)
Parallel signatures of sequence evolution among hearing genes in echolocating mammals: an emerging model of genetic convergence (2012)
Abstract: Parallel sites implicate functional convergence of the hearing gene prestin among echolocating mammals (2014)

Ed Yong, who should now be known as an excellent pop-science writer, has the story. (Though the title is misleading)
Ed Yong wrote:Based on the DNA sequences of these Prestin versions, Liu drew a mammal family tree (a ‘phylogeny’). It looked much like what you would expect, with the whales and bats clustering in separate family groups. But convert the sequences into amino acids and the picture changes dramatically. Suddenly, the family tree becomes utterly misleading. The echolocating mammals, be they bats or whales, are united as close relatives, to the exclusion of their rightful evolutionary kin.


Dandan wrote:I think you are misunderstanding what the authors are saying, when they talk about convergent evolution, they mean that the same genes evolved independently,( in other words convergent evolution at a genetic level,) this is not analogous to the alleged convergence between wings in birds and bats, where similarities are only superficial. In the case of the venom in platypuses and snakes we are talking about the same genes. Supposedly the genes evolved twice independently, like in the case of pristine in bats and dolphins the same genetic material evolved independently more than once, exactly what you said was impossible.


I'm sorry to say that in both cases, it is you who misunderstood the articles in question. In both cases, I went to great lengths to find out whether you were right or not. I contacted the authors, though of course none of them will reply. I read all the literature on the topic I could find, three and five papers respectively. In both cases, I BLASTed sequences to see if you were right. As a last resort, I asked for Rumraket's input and he agreed that I had correctly understood the papers, too!

In both cases, that is to say both in prestin and in the platypus venom, the proteins were remarkably similar, but they were not the same. And while they share some similarities, they are NOT the same, they are NOT identical. I can't stress this enough: There is great genetic difference between the two.

Again for those interested, here are the three (Open-Access) articles relevant for this discussion:
Defensins and the convergent evolution of platypus and reptile venom genes (2008)
Novel venom gene discovery in the platypus (2010)
Tracing Monotreme Venom Evolution in the Genomics Era (2014)

Dandan wrote:Both the bat and the platypus examples are examples of organisms evolving the same genetic material independently; exactly what you said was impossible. And just to be clear these are not isolated examples, there are thousands of similar cases.


As already stated, you're simply wrong. There are many examples of convergent evolution, where proteins and other structures are very similar, but I don't think there has been a single example where multiple genes have independently hit upon the exact same sequence.

Dandan wrote:The whole point I was trying to make is to show that it would be very unlikely for two unrelated clades to suffer from the exact same mutations, since we both agree with that conclusion I see no point in discussing on weather if my equations are correct or not


I already explained that your equations were wrong, even though your conclusion happened to be correct for once. How we arrive at a conclusion is at least as important as the conclusion itself.

Also, I did not give you permission to re-post what I sent to you in a PM. I would have agreed, had you asked, but this is rather bad form.

Dandan wrote:So basically you said that two unrelated clades can´t have shared mutations, and I did showed you examples of shared mutations among clades that are supposed to be unrelated. ¿can I claim vicory?


No, absolutely not. First, you'd have to show that two unrelated clades actually share independent mutations. You haven't done that.

Dandan wrote:In fact I do agree with you, but I don´t think humans and chimps have any shared mutations, they have traits in common, but this are not due to shared mutations, nor due to common ancestry, but rather by “common designer”

You see chimps and humans have similar needs (digest food, breathe, pump blood etc.) it makes sense to say that a designer would design similar systems to satisfy similar needs, and It would also make sense to say that a designer would create similar genes to design similar systems.

For example Microsoft Word and Microsoft Power Point have a similar spelling check system, and I am sure that this system is encoded by similar codes in both programs, after all if both systems need a spelling check system, why wouldn’t the designer use the same code in both of the systems?

Same thing with DNA, if humans and chimps need to breathe the same air, why wouldn’t a designer create them with a similar respiratory system with similar genes?

In my model similarities among chimps and humans are not due to shared mutations,


"Common design, common designer" or "common genes, common designer". This is extremely bad logic and it's trivial to show why.

----------------------------------------------------------------------------------------------------------

Now as I said, I had two things to explain to you. I'll be putting them both in a ShowMore box, just to make the post seem shorter.
ShowMore 1 explains why it's the differences that matter as well as the similarities.
ShowMore 2 explains why "Common design, common designer" or "common genes, common designer" fails.





I hope that I have successfully addressed all your arguments, though the sheer volume may have resulted in me missing one. If I have indeed failed to address something significant, please don't hesitate to point it out to me so I can finish the job. I'm currently reading another 4 papers and hope to address the "discordance-argument" next time. I felt that I should take the time to properly read everything I could on the subject, and anyway this post is already way too long.
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Mon Jul 14, 2014 12:38 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

As already stated, you're simply wrong. There are many examples of convergent evolution, where proteins and other structures are very similar, but I don't think there has been a single example where multiple genes have independently hit upon the exact same sequence.


True, pristine in bats and dolphins is not identical, however contrary to what evolution predicts pristine in dolphins and bats have a higher degree of homology than pristine in dolphins and whales, the implication of such discovery and from an evolutionary perspective is that bats and dolphins suffered from the exact same mutations, exactly what you said was impossible, and exactly what you said that would falsify evolution.

Inferno wrote:Yes, I absolutely agree with that conclusion. As I said in the post:

"What you actually calculated (and where I provided the correct calculations) is the probability and time for a mutation to arise in independent populations. That is to say, I would be really surprised to see two identical mutations in two non-related species.".


I apologize from quoting from a private message, but since you didn’t share any personal data, I thought it wouldn’t be inappropriate to do so, but I honestly apologize. But the point is that you are admitting that it would be surprising to find 2 identical mutations in 2 unrelated clades, and in this particular case I presented an example of 14 shared mutations.

Quote from the article: Molecular Evolution: Gene Convergence in Echolocating Mammals," http://www.umich.edu/~zhanglab/publicat ... rrBiol.pdf

dolphins and porpoises share at least 14 derived amino acid sites in prestin with echolocating bats.


So even though I didn’t present a mammal with feathers I did presented something fundamentally equivalent.

In my equations I presented the probability for two independent clades sharing a mutation or a protein or a portion of DNA. Since we both agree that it is very unlikely for two organisms to suffer from the same mutations independently I thought there was no point in explaining my equation, but since you think this is relevant I´ll try to clarify what I tried to show with my equation.

In order to calculate the probabilities of two organisms sharing the same mutation, you need to know 2 things.
1) How probable is it for a member of a population to suffer a mutation in a specific place.
2) How probable is it to achieve fixation in this particular mutation.


Number 1 can be measured this way:
(Mutation Rate )(Population size)/Genome Size

Number 2 can be measured this way.
1/(Population size)(2)

If you multiply number 1 and number 2 you will end up with the probability for an organism to share the exact same mutation than an unrelated organism.

We know that the Mutation rate is around 100 mutations per generation and we know that genomes are around 3 Billion base pairs long, and for your convenience we can assume a population size of 1 individual.

*A larger population would make this less probable.


So if we substitute values.
((100)(1)/3,000,0000,000)) x (1/2)

We have a result of 1/60,000,000 this is the probability of having 1 shared mutation, in the particular case fo perstine we require 14 shared mutations, so you have to multiply 1/60,000,000 x 1/14 = 840,000,000…840,000,000 is the probability of having such a degree of homology in perstine between bats and dolphins. Such an even would have had to happen in 50,000,000 years (after the ancestor of bats that use sonar diverged from those who doesn´t)

So you have 50,000,000 years for an event that has a probability of 1/ 840,000,000 I think it is safe to assume that it never happened. And all this assuming a population of 1 individual, imagine what would happen if I would have used a more realistic number like 1,000,000…

But as I said before you don’t ‘have to agree with my equations, because we both agree that the probabilities are very small.

So why isen´t perstine in bats and dolphins fundamentally equivalent from a mammal with feathers?

And please do not forget that I am presenting present just as an example to illustrate my point, I am not suggesting that this is the only discordance that has ever been found.

Talking specifically about echolocation there are around 200 genes that are more similar in dolphins and bats than among animals that are supposed to be close relatives. In other words you don’t ‘have to deal just with perstine but with 200 genes in a similar situation.

“Despite being separated by millions of years of evolution, dozens of genes in dolphins and bats changed in the same manner to give the species their ability to echolocate. A study, which has received criticism from some scientists, found that similar mutations allowed webs of genes in the animals to contribute to the superpower of being able to “see” with sound….. The researchers report September 4 in Nature that nearly 200 genetic regions stood out as evolving together – far more than the researchers had expected.”
https://www.sciencenews.org/article/man ... holocation



So why isn’t this web of 200 genetic regions with unexpected homology among distant relatives (dolphins and bats) equivalent to a mammal with feathers? You can ignore all the rest of my arguments if you what, but please answer unambiguously to this question ¿why isn’t homologous echolocation in bats and dolphins equivalent to feathers in mammals?




.
It further seems that the convergence is only regarding the phenotype, not the genotype. Now that would have been more remarkable. As I said before, prestin is a protein, so while we might expect the protein to remain the same, the amino acids have changed and the genes most certainly have. .


No sir, the converge is observed at a genotypic level, the article that you quoted talks first about the phenotypic convergence (as an introduction) and then it talks about the genotypic convergence.

.[quote from the article that you quoted]
However, in trees
based on amino acid sequences,
constructed using a range of different
phylogenetic methods, we found
that the echolocating dolphins now
formed a well-supported group
with echolocating horseshoe and Old World leaf-nosed bats
.[/quote]

I don´t what to sound mean, but it´s more than clear that you didn´t read the article, since the whole point that the author was making is that we do observe convergence at a genotypic level.

Predictions…
.
I didn't say anything about "98% similarity" now, did I? I should have phrased that better and will explain that now. There are three examples I want to highlight:
Other great apes have 48 chromosomes, while we have 46. If we had simply lost a chromosome during our evolution, we would probably not exist. So there absolutely must be evidence of chromosomal fusion, otherwise we can’t explain how this miracle happened. Lo and behold, Human Chromosome 2 shows evidence of fusion. If a designer had designed the chromosomes, why didn't he do the same thing as with the other great apes? That’s what I would expect of a designer... but this? It can only be explained by evolution... or by a dumb designer.

Back in 1870, Darwin predicted that we should look for our common human ancestor in Africa. The Out of Africa theory (or "recent single-origin hypothesis", if we’re to be correct) is now more or less universally accepted.

Third, “almost all animals make Vitamin C inside their bodies. It was predicted that humans are descended from creatures that could do this, and that we had lost this ability. (There was a loss-of-function mutation, which didn't matter because our high-fruit diet was rich in Vitamin C.) When human DNA was studied, scientists found a gene which is just like the Vitamin C gene in dogs and cats. However, our copy has been turned off.”
This video explains it rather well.
.

Chromosome fusion:
1 As far as I can tell that was not predicted by anyone, can you quote from a paper, and interview, an article or something where a scientists makes that prediction BEFORE the discovery.

2 I don’t ‘think there is strong evidence that the fusion ever occurred, all you have is something that looks more less like a telomere in the center of Chromosome 2 most chromosomes have a similar pattern.

3 Even if there was a fusion, as a creationists all I have to say that a HUMAN suffered from a fusion a few thousand years ago, I don´t need ancient ape-man creatures nor millions of years to explain it.

Ancestor in Africa
Even though it´s interesting, I think it would be another topic, unless you correct me this has nothing to do with phylogenetics

Vitamin C
I agree, data shows exactly what we would expect if evolution where true, this is a correct prediction, however as I mentioned earlier most genes contradict evolutionary predictions, this gene simply happened to be an exception. Apes have dozens of thousands of genes finding 1 gene that happens to correspond to evolutionary predictions is really not that amazing.

Familly trees…

I think I already said what I had to say, I don´t have anything else to add, you can organize all living things in a big tree, or organize them all in many small trees, you can do the same with robost, cars, computers and chocolates.

Image
http://img13.imageshack.us/img13/4683/c ... barevo.jpg

The fact that unrelated kinds have traits in common doesn’t automatically imply that they share a common ancestor; we know this because we do see unrelated chocolates, computers and robots that have characteristics in common.


I would like to make a small summery
1 You did said that a mammal with feathers would falsify evolution because that would require that this mammal suffered from the exact same mutations that birds did, which would be very unlikely.
2 I did presented something equivalent, echolocation in bats and whales, which would require that bats and dolphins suffered independently from the same mutations (we are talking about hundreds of mutations)


You can answer to the whole reply if you what, but it would be ok with me if you focus only on this 2 particular points.
Tue Jul 15, 2014 6:48 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

Dandan, if I could ask a favour: If you want to highlight things, use functions such as underline or bold font, but please don't make words twice their original size. It makes for extremely irritating reading.

With that minor nuisance out of the way, let's dive straight in. I'll be going back to my original style: Addressing an argument as a whole, not individual quotes.
I will address three arguments:
1) Prestin, platypus venom and convergence
2) Faulty equations
3) Hershey candy bar family trees and other fictional phylogenies

----------------------------------------------------------------------------------------------------------

In previous posts, you cited both the hearing protein prestin (in bats and cetaceans) and venom (in monotremes and snakes) as evidence that genetic sequences hit upon the exact same sequence in species where not only we wouldn't predict them to be, but even where they would falsify evolution. In order to prove that, you must show that at least one such example is indeed correct.

Ardent observers will note that you have already dropped the example of platypus/snake venom convergence. However, I want to make it absolutely clear that I was right and you were wrong. I contacted the primary author of the paper (Novel venom gene discovery in the platypus (2010)) in question: Dr. Camilla M Whittington.

The original message was sent on Wednesday, July 9th, containing (among other things) the following:
-----Original Message-----
From: Benedikt Lernhart [mailto:XXXXX]
Sent: Wednesday, July 09, 2014 2:47 AM
To: Camilla Whittington
Subject: Platypus/snake venom genes

...

In the first article, you report 83 novel platypus venom genes which express proteins that are functional analogues of the proteins found in snakes. At no point in the study does it mention that the genes are homologous or identical, is that correct? In fact, as far as I read it, the study didn't even mention a comparison of snake/platypus genes, the comparison was between venom proteins, correct?
The second study is further evidence that snake and monotreme venom evolved independently and, though they share similar characteristics, are quite different from one another.

I hope that I've correctly understood these aspects of the research and I also hope that I might contact you again with further questions regarding the subject, should your time permit it.

...


Her reply came on Monday, 14. July 2014, 05:19:15 a.m.

I have not got explicit permission to quote her answer, so I will wait until I have that permission expressis verbis. However, I will summarize her reply:
Dr. Whittington is upset to see that the work of evolutionary biologists is being distorted by creationists to support false arguments. She specifically said that the 2010 paper did not find identical genes in platypus and snakes, that in fact they had never looked at snake genes. Rather, they compared the protein sequences and found similarity. She explained that the 2014 paper shows that animals share venom characteristics, which is the result of using a similar ""toolkit" of proteins" (direct quote). At the end of the E-Mail, she suggested another article on the topic, though I haven't been able to read it yet.
Bryan G. Fry (2005) From genome to “venome”: Molecular origin and evolution of the snake venom proteome inferred from phylogenetic analysis of toxin sequences and related body proteins, Genome Res. 2005 15: 403-420

So tell me Dandan: Who was right, you or I? That's a rhetorical question by the way, I was right.
I want a promise from you, Dandan: Promise never to use that particular example in your arguments, because it's wrong. Will you do that, now that I have shown it to be wrong beyond the shadow of a doubt?



Onwards to prestin. I can do the exact same thing: Contact the authors, ask them who was right and give their reply. That seems to be a task in futility, because I will remain right once again. Do you really want me to humiliate you so badly? Here's a suggestion: You contact them. Ask them if you've correctly understood their papers and be prepared to accept that you're wrong when even they tell you so.

In the exact way Dr. Whittington explained that the amino acids converged in platypus and snakes, meaning the protein sequence, so also do the protein sequences in bats and toothed whales. This is what you get wrong and that's why your example is not valid. Let me make this absolutely clear: When comparing the gene or nucleotide sequence, bats and toothed whales are correctly classified. If the protein sequence is compared, bats are classified with toothed whales. That's the whole deal. The gene sequences are not the same. This is made absolutely clear in a paragraph I quoted earlier:
Convergent sequence evolution between echolocating bats and dolphins (2010) wrote:To test whether convergent changes in bat Prestin genes have also occurred in echolocating whales, we sequenced the entire gene in a range of echolocating toothed whales and non-echolocating baleen whales, as well as additional bats (see Table S1 in the Supplemental Data available on-line with this issue). Trees based on nucleotide alignments from this larger dataset strongly supported the accepted species tree topology, albeit with the clustering of echolocating bats reported earlier [3]. However, in trees based on amino acid sequences, constructed using a range of different phylogenetic methods, we found that the echolocating dolphins now formed a well-supported group with echolocating horseshoe and Old World leaf-nosed bats (node posterior probability = 0.99 or 0.94 depending on the analysis), members of which emit Doppler-sensitive signals dominated by a constant frequency (CF) component [6] ( Figure 1A). Intriguingly, the addition of the sperm whale, which appears to echolocate at much lower frequencies [7], was seen to decrease support for this convergent signal, leading to the cetaceans and bats both forming monophyletic groups. The extent of sequence convergence between bats and whales was thus not sufficient to unite these clades when non-dolphin odontocetes were included in the analysis.


Do you understand your mistake, can we move on?

It's even more fun that you accuse me of not reading the article when you make a grave mistake:
However, in trees based on amino acid sequences, constructed using a range of different phylogenetic methods, we found that the echolocating dolphins now formed a well-supported group with echolocating horseshoe and Old World leaf-nosed bats.


You quote the passage that disproves what you claim, yet you fail to register that. Reading comprehension fail if ever there was one.

Now as I said, you can have the same amino acid sequence even if the underlying genetic code is different (I even showed that using the BLASTed sequences) and you can have very similar amino acids forming basically identical proteins, which is what they found. Don't let the pop-science articles cloud your judgement on this.




With all this, I can (once again) answer your question: "Why is this fundamentally different from mammals with feathers?"
If there really were identical genetic sequences at a level unexpected, unpredicted and not supported by evolution, you would be right. However, that is not what we find. Instead, we find convergent evolution at the protein, possibly amino acid level, not at the nucleotide level. We find incomplete lineage sorting accounting for a lot of "discordance" (I'm using it the way you're using it, by the way) with closely related species. We find that evolution happens and that creationism is still wrong.

----------------------------------------------------------------------------------------------------------

I will continue with the second point, your equations. It is strange that the first time you used an equation, you claimed it was (Mt x Ps x Pt x Nb)/Gs, but that it is now (Mt x Ps)/Gs x 0.5
That's very odd, isn't it? One might almost think you're just throwing a bunch of numbers/terms together and multiplying and dividing them at will.

Let's start with something positive: You are absolutely correct that the probability of achieving fixation is defined as P=1/(Ps x 2), which we worked out to be 1:2,000,000. However, please note that this is the probability of achieving fixation for a mutation with no selective pressure! Zacky (2007) quotes the equation for that on page 5, equation 2. It's not relevant, just thought I'd mention that little tid-bit. That being said, this is the only positive thing I can say about your equations and it's only after I corrected you on it, so that's hardly a big feat.

As for the probability of having (not suffering!) a mutation in a specific place, Sean B. Carroll already gives us the equation in "The making of the fittest" (2006): The sites per gene (you said 15) x the genes per mouse (I used 1 to make it easier) x the mutations per 1 billion (you said 100). That's the probability of one person having the mutation, that's when population comes in. You now need to figure out the amount of offspring your original population leaves (which I gave as 10,000 per year, which is more than reasonable) and you get 5 mutations every 1000 years. Those are the numbers you get when you use the correct equations. This is markedly different from the two numbers you got: 18,000,000 and... well either 1/60,000,000 or 1/840,000,000, I'm not quite sure. The problem is that your equations are so wrong, it's hard to keep track of what you're trying to do.

You won't believe me and I advise you not to believe me. Never trust someone else, always repeat what the other person has done. That's what I've done with your equation: You claim that a more realistic population size (like 1,000,000) would make this event less probable. Fine, let's use that number then.

You originally claimed
((100)(1)/3,000,0000,000)) x (1/2)
so all we have to do is substitute 1 by 1,000,000 so the new equation looks like this:
((100)(1,000,000)/3,000,0000,000)) x (1/2)
Correct?

Here's an oopsie: When I calculate that, I get 1:60 or one mutation every 60 years. So following your advice, I multiply 1/60 x 1/14 and I get 1/840. Meaning we should hit the 14-amino-acid lottery every 840 years. Strange, didn't you say using a larger population size should give us a lower probability? That's funny, it almost seems like your "equation" doesn't calculate what you think it does.

Even worse, any sufficiently large enough population and small enough genome (think Archaea or Bacteria) is guaranteed to give probability values of higher than 100%. If you know anything about probability, then you know that shouldn't happen. Is it maybe because what you're calculating isn't a probability value, but rather the time it takes to achieve something, like Sean Carroll did? If that were the case, numbers greater than 1 make sense: You'd have 10 mutations a year/generation or 20 or 30 or 100. Face it, Dandan: You're not calculating what you think you are.

Here's a tip, dandan: Next time you want to play with the grown ups, make sure you use equations from proper scientific sources and don't just make stuff up. You're insulting our collective intellects by not even checking if what you say is internally consistent, let alone right.

This is why I don't trust your "equations": They're wrong. The ones I use are simplified, but they're used in the scientific literature and I'm fairly sure they're pretty accurate. Can you claim the same thing? No, you can't.

----------------------------------------------------------------------------------------------------------

The third example you bring is one of fictional phylogenies. You present a "candy bar family tree" as proof that all things can be organized in one way or another and that this doesn't show that they have a common ancestor. This is a similar argument another creationist tried four years ago with AronRa. He proposed an organization of clothing, but it's basically the same argument. Aron showed that the other guy was wrong four years ago, so I'll probably be able to show that you're wrong now. Wait, scratch that probable, I can show it.

Before looking at the tree, let's first remind ourselves that candy bars do not reproduce and that therefore there is no descent with modification. Big candy bars don't give rise to little candy bars and mixing dark and white chocolate doesn't produce milk chocolate. I think so far we can agree.

When grouping candy bars, there are many possible ways to group them.
Do I group them by producer? In that case, all Côte d'Or (CO) products are closely related to one another and all Lindt & Sprüngli (LS) products are closely related to one another but they're only distantly related to Côte d'Or. We'd have the common ancestor of all candy bars and them as offspring.


Candy bar ancestor __________ Lindt & Sprüngli
\_________ Côte d'Or

Or should we group them by cocoa content? In that case, all CO and LS products with a high cocoa content are grouped together and all CO + LS products with low content are grouped together.

In fact, we'd have paraphyletic groups:

______ LS high cocoa content
__/______ CO high cocoa content
Candy bar ancestor __/
\_________ CO low cocoa content
\_______ LS low cocoa content

And so on and so forth. I could group them by nut (or other ingredient) content, by colour, density, texture... whatever I choose. This is basically a Linnaean-type taxonomy, but I already explained in the first post:
He classified living things according to artificial parameters: For example, plants were classified according to how many stamen a plant had. This system might be practical when classifying or identifying plants, but Linné realized that other systems could be more natural and reflect their genealogy or ancestry: This system would be called "phylogenetic system".


A Linnaean taxonomy does not show true ancestry of things, but merely reflects our own biases and pre-conceived notions. That's why it's stupid to group living things that way and that's why we needed a new, unbiased system. That's what the phylogenetic system was to accomplish.

You agree that we can group or map our ancestors in an ancestry chart or family tree. You will also agree that there is only one possible way of grouping your ancestry: You are the son of your parents, who are the children of their parents and the siblings of their brothers and sisters and so on and so forth. There is only one way. I touched upon this before: My uncle looks more similar to me than my father does, but we'd still not group my uncle and me closely together to the exclusion of my father, that would be stupid. There is exactly one way my family tree can look and it's reflected in my DNA.



In exactly the same way, there is only one way life can be classified if evolution is correct (within the limitations explained about three times now). However, there are any number of ways a tree of life could look if a creator were responsible for like on earth. I explained this too, with the emergence of watches from auto-mobile manufacturers.

If there really is/was a creator who magically "poof"ed all life into existence, then life doesn't necessarily need to be created. Elephants could be closely related to mice and chimpanzees could be related to spiders. There wouldn't necessarily have to be a rule that creator must follow.

I can back this up very easily: Ferrari creates some of the most amazing auto-mobiles in the world. (Some may disagree, but that's hardly relevant.) Ferrari also creates watches. Are those watches more closely related to Citizen watches or to Ferrari auto-mobiles?
A creator can simply start afresh, use a new template and create something completely different, yet the vastly different objects would still have identical ancestors: Ferrari or God, depending on where we're at.

Why does this never happen in nature? Why is the phylogenetic tree we see only compatible with evolution and never compatible with creation?

----------------------------------------------------------------------------------------------------------

I hope that I achieved four things in this post:
1) Show that my original interpretation of both the prestin and venom situation were accurate and in accordance with the scientific literature.
2) Show that Dandan's equations are both internally inconsistent and incompatible with the scientific literature.
3) Show that creationism isn't compatible with the twin-nested hierarchy we see present in nature and that only evolution can account for that.
4) Give people the opportunity to have a few laughs. I was especially pleased with the way argument 2 played out. When I crunched the numbers, I thought: Hang on, I know Dandan's wrong but... on such a fundamental level?

----------------------------------------------------------------------------------------------------------

Finally Dandan, you ask two questions:
Dandan wrote:1 You did said that a mammal with feathers would falsify evolution because that would require that this mammal suffered from the exact same mutations that birds did, which would be very unlikely.


Agreed.

Dandan wrote:2 I did presented something equivalent, echolocation in bats and whales, which would require that bats and dolphins suffered independently from the same mutations (we are talking about hundreds of mutations)


Wrong, and fundamentally so. Both your equations and the examples you present are not evidence for what you claim they are. Instead, you completely misunderstood every paper and argument brought forth so far and made a few mistakes so glaring, I can't believe you didn't spot them yourself. I'll let you in on a little secret to save us both some time: There is not a single example where two independent populations hit upon the exact same genetic sequence to do something. We'd know about it because it'd be huge news.
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Sun Jul 20, 2014 12:06 am
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

Well at least we agree on something (quote from Inferno from a private conversation)

I was made aware that Rumraket contributed to this as well. He explained that if evolution is correct, we should see this
Image

If creation is correct or if evolution is wrong, we should see this:
Image


Yes I agree if evolution where true we would expect to find the first tree pattern and if evolution where wrong we would expect to find the second tree pattern.

The point that I am trying to make is that we actually do find the second pattern

For example in the second picture, the first circle (left to right) could represent bats and the last represent dolphins, the last line could represent echolocation. You can´t make echolocation fit in the first tree.

And obviously echolocation is just one of many examples

About echolocation in bats and doplphins
Systematic analyses of convergent sequence evolution in 805,053 amino acids within 2,326 orthologous coding gene sequences compared across 22 mammals (including four newly sequenced bat genomes) revealed signatures consistent with convergence in nearly 200 loci. Strong and significant support for convergence among bats and the bottlenose dolphin was seen in numerous genes linked to hearing or deafness, consistent with an involvement in echolocation. Unexpectedly, we also found convergence in many genes linked to vision: the convergent signal of many sensory genes was robustly correlated with the strength of natural selection
http://www.nature.com/nature/journal/v5 ... 12511.html


The author is clearly and unambiguously talking about convergence at a genetic level, this is not parallel to the “convergence” that we see in bird and bat wings where similarities are superfitial.

As an evolutionist you are forced to believe that dolphins and bats suffered form the exact same mutations 200 times, which as you already admitted it´s virtually impossible.

So how do you explain the fact that echolocation in bats and dolphins (two distant relatives) is so similar even at a genetic level?
Whatever your explanation is; wouldn’t you use the same excuse if a mammal with feathers is ever found

Do you understand your mistake, can we move on?

You can concentrate in minor details if you what, however it is still a fact that your model requires the assumption that bats and dolphins suffered from the same random mutations, something that you already admitted that it´s impossible.

About my ecuation...


I honestly don´t understand why should we concentrate in this, if we both agree with the conclusion (it is very unlikely for two independent clades suffer from the same mutations by chance)

Just to clarify:

Both of my equations are the same equation, the only difference is that the second time I divided the whole equation in 2 parts
1 the probability for an individual to suffer from a mutation in a specific place
2 the probability of fixiation
Also in my second equation I assumed a population of 1 individual rather than 1,000,000 I did it because I wanted to show that even in an unrealistic scenario.

You made a good point, mutations with selective benefit are more likely to become fixed, however in my opinion you make the problem even worst because the mutation rate for beneficial is much smaller.

But feel free to prove me wrong, provide your equations and prove that it is reasonably possible for two organisms to suffer from the exact same mutation 200 times, you can assume ether positive or neutral mutations, you can assume ether a big or a small populations etc. please provide your numbers.

About Venom in platypus
She specifically said that the 2010 paper did not find identical genes in platypus and snakes


Just ot be clear, I never said that snakes and platypus have identical genes, I said that the proteins responsible for the venom have a high degree of homology, which would require homologous and independent mutations in snakes and platypus.

But if Whittington explicitly told you that this statement is wrong, then I am wrong, I don´t have problems in admitting my mistakes.


My only conclusion is that if evolution where true, you are forced to believe that bats and dolphis had the same mutations independently, you will obviously disagree, but any neutral observer can read the paper.

The implications of these are:

-According to what we know, It is extremely unlikely

-If echolocation can evolve twice independently, so can feathers, (meaning that a mammal with feathers won´t falsify evolution)

-The tree that predominates it´s the second
Fri Jul 25, 2014 4:43 pm
InfernoContributorUser avatarPosts: 2298Joined: Thu Apr 30, 2009 7:36 pmLocation: Vienna, Austria Gender: Cake

Post Re: Debate: dandan vs Inferno

This will be my 8th post in this debate. I'm looking forward to finally ending this debate, both because I'm leaving the country in two weeks and because this is absolutely pointless. Every single point in the last post has been addressed at least once and I believe I've provided ample evidence that the creationist perspective on them is wrong.

Nevertheless, I still have these three posts before me. Note that I will quote again and that I have to move the quotes around a bit to make my points.

----------------------------------------------------------------------------------------------------------

Dandan wrote:The point that I am trying to make is that we actually do find the second pattern


You're absolutely wrong, which is what I've been trying to explain for the last seven posts. As every single paper both you and I quoted explain, the species tree (meaning the one built on multiple genes) reflects the first tree in the post. The few lines diverging from the species tree are the few examples where individual genes occur in more distantly related species, aka Incomplete Lineage Sorting.

For example in the second picture, the first circle (left to right) could represent bats and the last represent dolphins, the last line could represent echolocation. You can´t make echolocation fit in the first tree.


Again wrong. In the first picture, the first circle could represent a cow (cousin of dolphins), the second a dolphin (note close relationship to cows on species tree) and the third bats (note line going from dolphins to bats, representing echolocation). Of course, that's a strong oversimplification.

Dandan wrote:Just ot be clear, I never said that snakes and platypus have identical genes, I said that the proteins responsible for the venom have a high degree of homology, which would require homologous and independent mutations in snakes and platypus.


That's simply not true, you specifically stated that they have the same genes, which I have to understand means "identical".
I checked a few dictionaries just to make sure:
Merriam-Webster The same: being one without addition, change, or discontinuance : identical
Oxford Dictionaries (the same): 1 Identical; not different:
...and so on and so forth.

Dandan wrote:That is not true, platypuses and snakes have the same proteins (same genes) for venom, but that is irrelevant you will still invoke convergent evolution as you did with the sonar in whales and bats.

...

I think you are misunderstanding what the authors are saying, when they talk about convergent evolution, they mean that the same genes evolved independently


Will you admit that you definitely said "same genes" meaning "identical genes"? Will you also admit that you claim the same thing (same aka identical genes) for dolphins/bats?

Dandan wrote:The authors explain this by arguing that bats and dolphins received the exact same mutations independently, exactly what you said was impossible.

...

Supposedly the genes evolved twice independently, like in the case of pristine in bats and dolphins the same genetic material evolved independently more than once, exactly what you said was impossible.

...

The whole point I was trying to make is to show that it would be very unlikely for two unrelated clades to suffer from the exact same mutations, since we both agree with that conclusion I see no point in discussing on weather if my equations are correct or not

...

3)I did presented something equivalent, pristine in dolphins is more similar to bats than to whales, which would imply that dolphins suffered from the exact same mutations than bats independently


And so on and so forth. I was only half-way through our conversation at that point, so I assume there are more quotes where these four came from. This lets us deal with the following:

Dandan wrote:The author is clearly and unambiguously talking about convergence at a genetic level, this is not parallel to the “convergence” that we see in bird and bat wings where similarities are superfitial.


I already explained that the similarities are exactly analogous to the ones in the platypus/snake story: Very similar or identical proteins are coded for by a very different amino acid sequence. This isn't even up for debate!

I also explained that this is hardly surprising: Very similar living conditions (little light in caves and deep water, but also water as a medium for sound waves) favour similar solutions (echolocation) which require similar or identical proteins. It's hardly as surprising as you make it out to be.

Dandan wrote:As an evolutionist you are forced to believe that dolphins and bats suffered form the exact same mutations 200 times, which as you already admitted it´s virtually impossible.


I think I already said it, but "suffer" sounds like it's some horrible genetic experiment (á la Abomination from "Incredible Hulk") gone wrong. You understand that every human being on average "suffers" from 150 mutations at birth?

In any case, we've already established that the "exact same mutations" aren't actually what any of the previous papers said. Now you're chucking another one at me? I've read it and I'm not sure (I'm trapped in a lot of technical stuff) but I doubt it confirms your position. But I'll remain open to proof. Contact the authors and explain the situation. Ask them who's right, just like I did.

Dandan wrote:So how do you explain the fact that echolocation in bats and dolphins (two distant relatives) is so similar even at a genetic level?
Whatever your explanation is; wouldn’t you use the same excuse if a mammal with feathers is ever found


You do understand that similarities aren't a problem at all, that they're even predicted? I already explained that about three times above. What you need to find is not similarities in proteins, but identical nucleotide sequences where there shouldn't be any. Do you understand the differences between the two, yes or no? If you don't, I'll explain them to you so you understand what you're looking for.

Dandan wrote:You can concentrate in minor details if you what, however it is still a fact that your model requires the assumption that bats and dolphins suffered from the same random mutations, something that you already admitted that it´s impossible.


How is that a "minor detail"? I explicitly showed that you're wrong about "exact same genes", which was the crux of your argument. I was able to show that the similarities occur in the amino acid sequences, something that's perfectly understandable. I think I even explained at some point that "identical amino acids" in no way mean "identical nucleotide sequences". Do you even understand the difference between the two?

Codons code for amino acids, meaning three nucleotides in a row. Take the amino acid Leucine: TTA, TTG, CTT, CTA, CTG, CTC all code for it. If we have the amino acid sequence LLLLLL, we may have any of the six codons coding for them. Suppose that bats and dolphins both have the same amino acid sequence "LLLLLL". The bats may have the codon sequence CTCCTCCTCCTCCTCCTC, while dolphins may have the sequence TTATTATTATTATTATTA. Both code for the same amino acid sequence (LLLLLL), but they're not identical, would you agree? It would therefore be idiotic to claim that the two could only arise by the exact same mutations, would you agree?

Crap, I just noticed I put this in the post that was deleted. Never mind. Basically what you can do is go on BLAST @ NCBI and find the gene prestin. You can then compare the amino acid sequences in bats and dolphins. The two are not identical, but they're similar up to a point. In one of the two, the gene sequence is known, so you can copy that sequence and look for similar sequences. Tell me what you find, because when I did it, I found two very different gene sequences. Funny that, similar amino acid but different nucleotide sequences. Isn't that something I said since one of the first posts?

But what you'd need for your claim to be true would be two identical nucleotide sequences. Where are they? Why have you claimed that they exist, but have yet to show them to me?

Dandan wrote:Both of my equations are the same equation, the only difference is that the second time I divided the whole equation in 2 parts


No they're not. You can't just insert an equation where you want to, you have to actually replace a term.

As I explained, your original one is: (I erroneously wrote Pt while it should have been Pf in the first post.)
(Mt x Ps x Pf x Nb)/Gs

Your new one is
(Mt x Ps)/Gs x 0.5

That's something completely different. Where did the term "Pf" go? I know exactly what you did, but you made a critical mistake. In the second (mathematically correct) one, you had:
(Mt x Pf)/Gs x 0.5
and
Pf = 1/2Ps
You correctly inserted the equation into the first and got (Mt x 1/2 x Ps)/Gs which is also (Mt x Pf)/Gs x 0.5
So far, you're right.

But in the first one, the term "Pf" is there, while 0.5 is missing! That's certainly not correct. That's not mathematics you're doing here, that's number-juggling. Or equation-juggling, however you want it.

Dandan wrote:Also in my second equation I assumed a population of 1 individual rather than 1,000,000 I did it because I wanted to show that even in an unrealistic scenario.


You failed miserably. Did you read the part where I actually used a population of 1,000,000 and it actually proved my position?

Inferno wrote:You originally claimed
((100)(1)/3,000,0000,000)) x (1/2)
so all we have to do is substitute 1 by 1,000,000 so the new equation looks like this:
((100)(1,000,000)/3,000,0000,000)) x (1/2)
Correct?

Here's an oopsie: When I calculate that, I get 1:60 or one mutation every 60 years. So following your advice, I multiply 1/60 x 1/14 and I get 1/840. Meaning we should hit the 14-amino-acid lottery every 840 years. Strange, didn't you say using a larger population size should give us a lower probability? That's funny, it almost seems like your "equation" doesn't calculate what you think it does.


Do you admit that your equation is simply wrong, that it's never been used in academia? Do you further concede that the ones I provided in this post are the correct ones?

Dandan wrote:You made a good point, mutations with selective benefit are more likely to become fixed, however in my opinion you make the problem even worst because the mutation rate for beneficial is much smaller.


Really? How do you figure that?
You do understand that every loci has on average the same probability of being subject to a mutation?

Dandan wrote:But feel free to prove me wrong, provide your equations and prove that it is reasonably possible for two organisms to suffer from the exact same mutation 200 times, you can assume ether positive or neutral mutations, you can assume ether a big or a small populations etc. please provide your numbers.


Did you even read what I wrote? I sincerely doubt it, otherwise you'd not claim that I said this or in any way assumed it.

Dandan wrote:My only conclusion is that if evolution where true, you are forced to believe that bats and dolphis had the same mutations independently, you will obviously disagree, but any neutral observer can read the paper.


You already agreed that what we see is similarity in the amino acid sequence, correct? You understand that this is not the same as "identical mutations", right? You understand that "similarity in the amino acid sequence" poses absolutely no problem to evolution, that only "identical mutations" would? If you understand these three points, do you also understand why your point is wrong?

Dandan wrote:If echolocation can evolve twice independently, so can feathers, (meaning that a mammal with feathers won´t falsify evolution)


No, because the echolocation of dolphins and bats is based on the similarities in their amino acid sequences, while I stated that the feathers had to be identical at the nucleotide sequence. You understand the difference, right? However, we might evolve different feathers, maybe because our hair thickens, grows outwards and straightens. They might then be feathers, but the underlying structure isn't the same. Do you understand the difference?

Dandan wrote:The tree that predominates it´s the second


How come that every single scientist claims that the first tree predominates? The very papers you cite agree with me on that!




EDIT:

Doctor Whittington has yet to reply to my second E-Mail, but I believe I made it sufficiently clear in my first E-Mail that I wanted to quote her verbatim. So without further ado, here's her E-Mail.

Camilla Whittington wrote:Dear Benedikt,

Thanks for your email and your interest in our work. It's upsetting that the work of many evolutionary biologists is being distorted to support creationist arguments. You're right that in our 2010 paper we did not find identical genes in platypuses and snakes- we did not look at snake genes at all. Rather we see similarity in the venom proteins and this similarity is independently derived i.e. it is a result of convergent evolution. It is not indicative of common venom ancestry. The 2014 paper, as you say, shows that while the characteristics of venom are shared across different animals, this is the result of the recruitment of a similar 'toolkit' of proteins.

The venom of the platypus and snakes have evolved independently. The venoms show similarity because there are certain gene types that have characteristics that cause them to be preferentially selected during venom evolution. These characteristics might be things like existing expression in the tissue that evolves to become the venom gland. For example, beta-defensins are expressed in skin; this could be why the OvDLPs that are derived from beta-defensins are expressed in the platypus venom gland, because this venom gland is thought to have been derived from a skin (sweat) gland. In a similar manner, the snake venom glands are derived from salivary glands and there are lots of examples of snake venom proteins that are derived from genes with original salivary gland expression. Other characteristics that might cause genes to be preferentially selected for expression in venom glands are certain confirmations (bonds between cysteine amino acids, for example, are very important in maintaining protein structure and function), genes from multigene families, and the list goes on. You might be interested to read Bryan Fry's paper on the subject (http://genome.cshlp.org/content/15/3/403.full). His work on venom and evolution is really interesting.

Best wishes,
Camilla
"Sometimes people don't want to hear the truth because they don't want their illusions destroyed." ― Friedrich Nietzsche

"I shall achieve my objectives through the power... of Science!" --LessWrong
Sun Jul 27, 2014 3:20 pm
dandanPosts: 431Joined: Sat May 10, 2014 2:16 pm Gender: Male

Post Re: Debate: dandan vs Inferno

That's simply not true, you specifically stated that they have the same genes, which I have to understand means "identical".


No same genes doesn´t mean identical genes, for example humans, chimps and all mammals have the FOXP2 gene, it is said to be the “same gene” but the gene is not identical in all species, if you compare the FOXP2 gene in humans and chimps you will find some differences.

An analogy would be the bible (King James version) and the bible (new international version) it is said to be the same book, but it is not identical.


Obviously since we are suppose to share a common ancestor with chimps, originally chimps and humans where the same animal and therefore we had the same FOXP2 gene, according to evolution, the differences between the human FOXP2 gene and the Chimp FOXP2 gene evolved after they diverged. ¿Agree?

Obviously if we compare the human FOXP2 gene with the Cow FOXP2 gene we would expect to find even more differences because we diverged from cows longer ago than we did from chimps ¿Agree?

From an evolutionary perspective if humans and cows have a common mutation in the FOXP2 gene, that is absent in other primates, evolutionists would say that cows and humans received the same mutation independently ¿agree?
About echolocation.

This is what happened; scientists study a series of genes that are related to echolocation, these genes are present in all mammals however only a particular variant of these genes allows echolocation.

The problem is that bats and dolphins being far relatives, have the same variation at least in 200 sites, (closer relatives don´t have these variatons) the implication is that bats and dolphins received the same mutations 200 times independently.

So unless you prove that it is statistically possible and reasonable for 2 independent clades to suffer from the same mutations 200 times independently, I think you are forces to admit that I least this is a valid argument against evolution.





Of course I am aware of the fact that you don´t accept my “same 200 mutations” premise and you already explained your reasons, you don´t have to repeat the same information, but the papers are available for free and anyone could confirm that I am interpreting the paper correctly. your model requires that dolphins and bats suffered from the same mutations.


About Venom in platypuss
It seems to be a similar case as it is with echolocation, I emailed the author but she told me that she was out of town and that was going to answer until mid-august.

No they're not. You can't just insert an equation where you want to, you have to actually replace a term.

As I explained, your original one is: (I erroneously wrote Pt while it should have been Pf in the first post.)
(Mt x Ps x Pf x Nb)/Gs

Your new one is
(Mt x Ps)/Gs x 0.5

That's something completely different. Where did the term "Pf" go? I know exactly what you did, but you made a critical mistake. In the second (mathematically correct) one, you had:
(Mt x Pf)/Gs x 0.5
and
Pf = 1/2Ps
You correctly inserted the equation into the first and got (Mt x 1/2 x Ps)/Gs which is also (Mt x Pf)/Gs x 0.5
So far, you're right.

But in the first one, the term "Pf" is there, while 0.5 is missing! That's certainly not correct. That's not mathematics you're doing here, that's number-juggling. Or equation-juggling, however you want it
.

PF = 0.5 in a hypothetical population of 1 individual, with that clarified, do you agree that both equations are the same one?
The equations are supposed to measure the probabilities for two independent clades to suffer from the same mutations, not time as you seem to believe.


If you what to use a population size of 1 million you would have to change 0.5 for 1/2,000,000

However we both agree that it would be impossible for 2 independent clades to suffer from the same 200 mutations, so why are you still making a big deal with the equation?
Mon Aug 04, 2014 1:42 am
Next
Post new topic   Page 1 of 2
 [ 26 posts ] 
Return to Debates Archive

Who is online

Users browsing this forum: No registered users and 0 guests