Testing the creationist hypothesis

[T_aquaticus]

@dsterncardinale

If the creationist “orchard of life” is correct, we can expect nested hierarchies on constrained sequences for all cellular life, but we should only see nested hierarchies for unconstrained sequences within each “kind” and no correlation “between kinds”. We actually see strong correlations on constrained and unconstrained sequences, which invalidates the separate ancestry hypothesis.

As discussed in “Faith and the Human Genome” by Francis Collins, the difference in divergence across exons and introns is rather inexplicable with separately created kinds (i.e. the orchard of life). Creationists tell us that it makes sense for God to copy a gene, and then tweak it so it functions a bit differently in a new species. That makes sense for exons. However, why do we see more differences in the introns than in the exons? That makes no sense within creationism, but it makes perfect sense for evolution and common ancestry. I snapped this picture of CAPZA2 at the UCSC genome browser, the same gene that Dr. Collins discusses in his essay.

It is not just a human/mouse comparison one can do. Eric Green at the Genome Institute has looked at this same region in many other species and, in fact, you can find this same CAPZA2 gene in everything from chimps down to zebra fishes and a lot of things in between (see Figure 4). Notice the pattern. The chimpanzee is almost 100% identical to the human, except the chimp has a deletion just before exon 2 that we do not have. Otherwise the match-up, as in most cases of human and chimp comparison, is about 98.5% to 99%. You can see that the baboon is starting to diverge. The cat and the dog and the cow all look a lot alike, and again if you look at the CAPZA2 exons, you will see that every one of those species has a nice conserved little segment there. But as you get further away to rats, mouse, chicken, two different kinds of pufferfish and then a zebra fish, about the only thing you see is the protein encoding regions, while the rest of the scattered noise goes away. Again, this is a very compelling kind of pattern in terms of what one would expect from evolution.
https://www.asa3.org/ASA/PSCF/2003/PSCF9-03Collins.pdf

[Elliot_Mudd]

@John_Harshman

Not sure what that has to do with the original claim. I take “constrained sequences” to be those under selection, and they are supposed to display a nested hierarchy throughout all cellular life, not just within dogs.

Ah right. Sorry I misunderstood your original point when I wrote this

@John_Harshman

But if the quite different idea you present is true, that’s great news for creationists, because we can now identify “kinds” easily: they end at the point at which nested hierarchy stops. Sadly, nobody can find such a point.

Yeah. To be consistent they’d have to say that humans and other primates are in the same “kind”. Which they obviously can’t.

[Elliot_Mudd]

@John_Harshman

But the point is that spermatogenesis goes on throughout a man’s life, so the older the man, the more mutations in the germ line. Women, however, have all the eggs they’re going to have at birth.

Right, but even then that wouldn’t create enough genetic variation in one mans offspring to explain how we can all descend from one man and one woman 6,000 years ago as creationist claim.

[dsterncardinale]

@John_Harshman

Why would that be?

Same designer, similar function, similar parts. “Engineer reusing parts” kind of thing. That’s how creationists explain universal nested hierarchies in things like ribosomes, for example. Doesn’t work for unconstrained regions, though.

[John_Harshman]

@dsterncardinale

Doesn’t work for unconstrained regions, though.

It doesn’t work for functional bits either. There’s no reason to expect functions to follow a nested hierarchy, and certainly no reason to expect different functional parts to follow the same one.

[Dan_Eastwood]

For many Creationist and Intelligent Design hypotheses you need look no farther than the sort of evidence they demand for evolution. For instance, Creationist may demand a complete fossil record, but it would be far easier to demonstrate that fossils do not fit the nested hierarchy expected for Common descent. For ID, Behe’s demand for a “Stepwise evolutionary pathway” to disprove design of irreducibly complex structures would not disprove Design at all, because that could always be the pathway the Design choose to use. Both of these examples depend on a misunderstanding of what it means to falsify a hypothesis.

[Rumraket (Mikkel R.)]

@dsterncardinale

Same designer, similar function, similar parts. “Engineer reusing parts” kind of thing.

That gives you similarity, not necessarily a tree, much less does it give consistently very similar trees from different loci. Again, why so similar trees from completely unrelated parts? Ribosomes might require tRNAs to interact with (and there certainly are direct interactions between different ribosomal components), but the constraints on their function are not towards tree topology, it’s constraint towards things like folding and function. An enormous number of alternative sequences are able to fold and function equivalently, so… why so similar trees?

@dsterncardinale

That’s how creationists explain universal nested hierarchies in things like ribosomes, for example.

That is the explanation they offer, but it doesn’t work. It does not explain why we should get nesting clades, as opposed to just a star phylogeny.

[dsterncardinale]

Yes, it’s not a great hypothesis, but I’m articulating the creationist position - the expectation that functional sequences should be similar in a nest hierarchical pattern. They claim that as consistent with independent creation events. Which, fine. We can test that, as described above.

[John_Harshman]

@dsterncardinale

Yes, it’s not a great hypothesis, but I’m articulating the creationist position - the expectation that functional sequences should be similar in a nest hierarchical pattern.

You articulate it much more than they ever do. They usually just wave their hands in the direction of “common design”.

@dsterncardinale

We can test that, as described above.

That runs into the problem that creationists commonly don’t believe that there’s such a thing as junk DNA. It’s all functional. And a lot of the functional differences aren’t even genetic, though what they are instead is seldom explained.

[dsterncardinale]

@John_Harshman

@dsterncardinale

We can test that, as described above.

That runs into the problem that creationists commonly don’t believe that there’s such a thing as junk DNA. It’s all functional. And a lot of the functional differences aren’t even genetic, though what they are instead is seldom explained.

That is 100% the default response, but it’s trivially easy to show that that’s incorrect. Heck, their own sources (ie ENCODE) show that they’re wrong. Not that a dedicated YEC will care that the supposed refutation is obviously false.

[T_aquaticus]

@Elliot_Mudd

Yeah. To be consistent they’d have to say that humans and other primates are in the same “kind”. Which they obviously can’t.

It gets even more interesting when they claim that apes are a created kind, excluding humans. As it turns out, there are more genetic differences between the species within their created kind than there is between chimps and humans. I would also hazard a guess that for many of the proposed created kinds that the divergence between the genomes within the kinds is more than the divergence between chimps and humans. And yet, they claim evolution could not span the gap between species like chimps and humans.

[T_aquaticus]

@Rumraket

That gives you similarity , not necessarily a tree, much less does it give consistently very similar trees from different loci. Again, why so similar trees from completely unrelated parts? Ribosomes might require tRNAs to interact with (and there certainly are direct interactions between different ribosomal components), but the constraints on their function are not towards tree topology , it’s constraint towards things like folding and function. An enormous number of alternative sequences are able to fold and function equivalently, so… why so similar trees?

MicroRNA’s might be a good example of what you are talking about.

The stem-loop structure is needed for cleavage by Dicer into the mature microRNA. The stem of the structure contains the seed region that binds to the target region on the 3’ UTR of mRNA transcripts, so the stem tends to have conserved sequence. However, the loop structure just needs to be non-complementary bases in order to form the loop, and there tends to be much less conserved sequence in that region. For example, here is a crude alignment of mir-219 across many species, with human sequence at the bottom and increasing evolutionary distance as you move up. By the time you get to distantly related species the only similar sequence left is the seed region and its complementary sequence at the other end of the unprocessed miRNA.

                        ********************* *                      *** *** *  ***** **                              
                   *    ********************* *               *    * *** *** * ****** **                              
                   **   ********************* *               *   ** *** *** * ****** **                              
                   **   ********************* * *             *   ** *** *** * ****** **    ** *                      
           *       ** * ********************* * * *        *  *   ****** *** ******** **    ** *                      
           *** *   **** ********************* * * *  *     *  *   ****** *** ******** **    ** *    * *               
           *****   **** *********************** * ** *     ** *   ********** ******** *** * ** *  * * **              
         ******** ***** *********************** * ** *    *****  *********** ******** *** ******  *******             
         ************** **************************** *    ***** ************ *******************  *******             
        *************** **************************** *    ***** ************ ******************** ******* ** ***      
        *************** **************************** *    ***** ************ ************************************     
      * ******************************************** *    ****************** *************************************    
     *********************************************** *    ********************************************************    
     *************************************************    ********************************************************    
     *************************************************    ********************************************************    
    **************************************************    ********************************************************    
    **************************************************    ********************************************************    
    **************************************************    ********************************************************    
    **************************************************    ********************************************************    
    **************************************************    ******************************************************** ** 
    **************************************************    ******************************************************** ***
    **************************************************    ******************************************************** ***
    **************************************************    ************************************************************
    ************************************************** * *************************************************************
    ******************************************************************************************************************
    ******************************************************************************************************************

Added in edit: in case anyone wants to check out the miRNA database I used for the figure above.

microRNAViewer