As John has pointed out, organisms inherit traits, they do not re-use or swap out components. There is no object oriented programming in nature, and no Motorola part numbers. The latest gear in life cannot be rolled out to all interested creatures on distant phylogenic branches. So your suggestion that Rum is force fitting evidence is ironic, because the whole idea of components is a feature of human manufacturing but an anthropomorphism in nature.
2 Likes
Letâs do this again. Here is a scientific paper featuring portions of a single orthologous gene in various species of the taxon Crocodylia. The gene is called C-myc, and the sequences include part of one intron, the following exon, and the 3â untranslated region of the gene. No two species have the same sequence for any of those bits, and the sequences of all three bits have the same nested hierarchy of differences. Can you conclude anything from that paper? (I believe youâve seen that paper before.)
3 Likes
Hi Bill,
Science certainly does have some limits. In particular, it cannot speak to any divine causality or purpose that might stand outside the normal operations of the universe. I believe that God did create the universe, that God has made self-revelation to prophets and above all in a Nazarean carpenter turned rabbi, and that God upholds even the ânormalâ operations of the universe by the mighty power of what is sometimes called providence. None of my beliefs are based in science, but in other ways of knowing.
That said, science is extremely useful for understanding those normal, God-ordained operations of the universe. That you stand so doggedly against the work of scientists working in biology mystifies me; why should we not welcome the work of those who think Godâs thoughts after Him, to paraphrase Johannes Kepler?
Neil Shubin, an evolutionary biologist, has explained the relationship between genetic mechanisms and major transitions quite brilliantly. His work was described to you in this post, which was accompanied by a link to detailed information.
If you possess some evidence of why Shubinâs (and many othersâ) work on HOX genes is wrong, by all means please share this evidence. And you might win the Finlay Prize while youâre at it. Otherwise, I am puzzled that you would repeat such an ill-informed statement about the work of biologists.
Scientists need a model with testable predictions. I am not aware of any model of separate ancestry that makes quantifiable, testable predictions. Are you?
In the meanwhile, statistical tests have been used to ascertain which genes were present in the LUCA population:
Nothing is stopping a genomic comparison of common ancestry (CA) vs. separate ancestry (SA) models other than the lack of a suitable SA model.
Given a suitable SA model, statistical methods can be applied to genomic data. A very interesting, peer-reviewed paper compared common ancestry of primates to three distinct SA models, and found that the CA model was astronomically more likely, given primate genomic data.
https://onlinelibrary.wiley.com/doi/full/10.1111/evo.12934
If you cannot access this publication, the free prepub is available on BioArxiv:
There is only one thing preventing this technique from being applied more broadly: the lack of an SA model that makes sufficiently detailed predictions as to be testable with genomic sequence data.
Respectfully,
Chris Falter
5 Likes
Hi Chris
I donât think the current data supports a single point of origin. How many points of origin? I donât know. If I am right the current models are misleading research. I hear additional ORF gene data is coming rapidly and hopefully it will help settle the issue.
I am currently looking at the data John posted and it does support a tree for the Croc family in the way he describes Does this mean all crocâs share a common ancestor? What I also see are lots of mutations fixed in the populations in a very limited space in the genome. Given these species( assuming ancestry) are closely related I am not sure how this could happen. A lot of the mutations are in introns which may very well be neutral. This may be fine depending time since the splits and the mutation rate but it will not be investigated if common descent is assumed base on the partial nested hierarchy inference.
Hi Bill,
I donât think youâve looked at the current data.
Good. Do you see the importance of nested hierarchy?
It was predicted by all crocs sharing a common ancestor before the sequences were in hand. Do you see the difference?
I think that your concepts of fixation and space are at odds with reality.
Mutation, selection, and drift. The math worksâyou should try it.
They almost certainly are.
Why donât you do the math right here?
Again, common descent is not assumed. It predicts the pattern, which is an observation, not an inference. Please stop misrepresenting the fundamental concepts of science.
6 Likes
This is good.
The strength of the tree support effectively rules out the separate ancestry hypothesis. How would it not?
The alternative to nested hierarchy â no relationship â was effectively ruled out by the likelihood analysis. So why would scientists investigate it further?
Geological and astronomical data strongly rule out the notion that the earth is flat. Why would scientists continue to explore the hypothesis that the world is flat?
Yours,
Chris
2 Likes
Hi Chris
I agree it is evidence for common descent but not strong enough alone to rule out separate starting points.
With the separate tree hypothesis the tree support can come from timing of the introduction assuming that the initial MYC gene is the same sequence design. From other data I have looked at the exons (coding regions) appear to be the same starting points. One area to investigate is the indels as they may effect gene expression timing if they are more than a few base pairs long.
@Winston_Ewert showed that design data can fit a tree. You cannot rule out independent trees based on a partial nested hierarchy alone.
This is a much simpler problem then sorting out the tree(s) of life.
One good future tool that can help here is getting accurate as possible pedigree mutation rate data.
Best
Bill
Bill, will you never tire of word salad? Nothing in that paragraph made the slightest sense.
He showed no such thing. Did you notice, incidentally, that both the 3â UTR and the indel data fit the tree perfectly with no room for any âmodulesâ?
1 Like
You have forgotten:
Ewert conceded that his methodology CANNOT BE USED ON SEQUENTIAL DATA
The crocodile analysis is a study of sequential data. Ewertâs method is therefore entirely unsuitable and inapplicable.
Bill, you appear not to realize how badly you misunderstand this subject.
I am pretty sure you understand the concept of statistical significance. Did the croc study authors not apply it correctly?
If they applied it correctly in ruling out the possibility of separate ancestry, why do you make the claim that they have not ruled out âseparate starting pointsâ?
I feel like a character in a Thomas Stoppard play.
Best,
Chris
1 Like
Ewert showed design can create a tree pattern with software module data along with gene module data. The programs with the most similar modules will likely have the most similar sequences. Does this mean the sequences could not also create a tree pattern given common design?
In Johnâs paper we are simply looking at part of a single gene. Have you read his paper yet?
They are not comparing against separate ancestry. They are comparing against no tree as a null hypothesis.
He did not really. We still do not know how well-supported or consistent that tree is, nor how including more (and different) programs in his analysis could alter his conclusions. It is entirely possible that he has an incidental tree in a very small data set, partially explained by some degree of real genealogy in the history of how the different java programs were created, that completely disappears once he includes more programs.
Programs donât have sequences. And in fact the modules are identical between programs, not similar. More word salad.
True, but what is your point? Do you claim that other genes would not give the same result? You are wrong. Both the studies mentioned in that paper and much subsequent work sample many other genes, with the same result. And the dependency graph theory is not compatible with, and is incapable of explaining, nested hierarchy of single genes.
Separate ancestry and no tree are the same thing.
If you show that design will not create a tree sure. The evidence with Winstonâs work and other discussions does not support this assertion.
Computer modules at their core are ones and zeros or in hardware differential voltages. If two separate programs share the same module the oneâs and zeros in those modules would most likely align. This is similar to the sequence of nucleotides.
Prove it. Itâs word salad, so I doubt you even know what you would be trying to prove. But please do try.
More complete word salad. Also false, as non-binary and even mechanical and analogue computers are possible. They have to function neither in binary (ones and zeroes) nor with âdifferential voltagesâ.
Why do you just blather out stuff like this?
There is no way to show that, since design could do anything. The question is whether thereâs any justification for supposing that design will create a tree. Nothing you have said implies any such justification.
No, the ones and zeros would in fact be exactly the same. And that isnât what we see in life, in which the DNA sequences analogous to those ones and zeros are not identical but differ in a nested hierarchy. That isnât compatible with any model of separate creation other than a deceptive attempt to similate common descent.
3 Likes
No. Under separate ancestry, there would be no expectation of shared mutations between intermediate nodes, therefore no expectation of a tree.
I cannot think of a clearer way to express this. Does it make sense to you?
If the English language has any meaning, then âseparate ancestryâ is equivalent to âno treeâ
Category error. Under a modules-based model, there is no measure of similarity. There is only a Boolean identity measure between two modules.
If you want to incorporate measures of similarity and distance, you will have to abandon Ewertâs approach, which cannot use similarity/distance measures.
Best,
Chris
3 Likes
I repeat for the nth time:
Ewertâs methodology cannot be applied to sequential data; therefore, it cannot be applied to the sequential data that are used in biology tree-building studies!!
Ewertâs paper built trees from sets of non-sequential, boolean variables (presence vs. absence of a module). His tree-building was based on distance in a very large dimensional boolean space, where each dimension corresponds with one module.
Biological research builds trees from sequential data (amino acid sequences or DNA sequences).
The math is different in critical respects.
Best,
Chris
3 Likes
If a module with a slightly different version as in a new version release you would see similarities and differences.
I agree if you find something you can positively isolate as a shared mutation then that is strong evidence of common descent.
There could be an expectation of a shared sequence. In order to know what is a shared sequence and what is a shared mutation you need to know the starting point.
Category error. Under a modules-based model, there is no measure of similarity. There is only a Boolean identity measure between two modules.
This is interesting. Let me do some homework here.