I still have no idea what point you were imagining you were making.
As it must, since it produces an exact fit. But the number of modules rises much faster than the number of taxa. Despite his claims, the model is overparameterized. Further, the graph, unlike a phylogenetic tree, has no explanation; that is, it doesn’t correspond to any real phenomenon.
More word salad.
More word salad. The gene pattern is explained by the tree. The tree is explained by common descent. The dependency graph, on the other hand, isn’t explained by anything.
Yes it does. The tree eliminates separate creation of the species on the tree. And you have no clue what the different kinds are or how to recognize them, when if there were such things they ought to be obvious.
Of course it isn’t. If what we’re talking about has nothing to do with science and is only about what a literal interpretation of Genesis would be, I don’t see much to talk about or any point to talking about it. Why are you at this late date bringing it up?
You don’t because you won’t look at the data and you have no criterion for declaring kinds. If you actually look, all four of those species are vertebrates, and vertebrates form a single nested hierarchy. Thus all vertebrates belong to one kind. What’s your evidence that they don’t? And please, for once, try to understand the difference between common descent and the origin of novelty. We’re talking only about the former.
There is no good explanation under the constraint of methodological naturalism. If I cannot discuss novelty based on your criteria then yes I would also conclude they are all one kind.
Do I think this is the right interpretation of all available data? I don’t as the observed nested hierarchy is not sufficient to rule out special creation. If the gene sets in Sal’s flower lined up and followed the branching pattern then I think there would be a good case for a single vertebrate kind.
There are fundamental problems with the theory and the rules of analysis that is leading us in the wrong direction. I think science needs to rethink the fundamental approach and a start may be a competing theory such as @Winston_Ewert has proposed. From our discussion I believe it can open up new ideas on how to analyze the data.
If indeed vertebrates do not share a common ancestor then the current line of thinking which discounts all competing theories is taking science in the wrong direction.
Please try to be clearer. No good explanation for what?
But they do line up and follow the branching pattern, as I have shown you several times. The exceptions are very few.
What are those fundamental problems? What is this competing theory?
That’s nice, but all evidence shows that vertebrates do share a common ancestor, and you have never managed to confront or even notice that evidence. How can we possibly proceed until you do?
Science doesn’t discount all competing theories. It only discounts those which have zero supporting positive evidence like Magic and Intelligent Design.
Why should your opinion matter to the scientific community? If you can’t demonstrate that an interpretation is wrong then why should we pay credence to your feelings about it being wrong?
Then show that they don’t, including gene deletions and gene gain.
We need evidence, not opinion.
You could say that about every single theory in science. If germs don’t cause infectious disease then germ theory is leading us in the wrong direction. If matter isn’t made of atoms then atom theory is leading us in the wrong direction.
What you keep forgetting is that you first need to show that the current paradigm is wrong, which you haven’t done.
There is not a good explanation for the pattern. The nested hierarch does not explain it as there are too many exceptions to the branching pattern. Only half the human genes are shared with all vertebrates in the Venn diagram. If the number was 98% and your explanation needed rigor on how these differences occurred it would still be difficult to reconcile the de novo genes and genes that did not follow the pattern.
This is an assertion that is not taking a realistic look at the data.
The competing theories are Nathaniel Jeanson’s, Winston Ewerts, and the RTB theory that @Meerkat_SK5 is presenting. Seeded animal populations described in Genesis 1 is the working hypothesis of these theories.
I agree under methodological naturalism which restricts the contradictory evidence. Methodological naturalism is leading to poor analysis in this case. @pnelson has tried to make this point to the group for several years.
I don’t think repeating this assertion is getting you anywhere when the data you have supporting the hypothesis has so many contradictions.
This statement needs evidence. How much noise would be expected? Are you factoring in gene loss and gain?
Why is this a problem for a nested hierarchy? A nested hierarchy will have synapomorphies and apomorphies, so it is expected to find genes that aren’t shared across the entire phylogeny.
So you admit that the evidence you have is not empirical and not observable? If so, why call it evidence?
You misunderstand what the pattern is. For what seems like the thousandth time, a pattern of presence and absence is treelike if it can be explained by only a single change on a particular tree. Genes present in all species need no explanation at all. And as you have been shown, the overwhelming majority of gene patterns can be explained by a single change, gain or loss, on the standard tree. There is no viable explanation for that pattern than common descent. For the few genes that don’t show a single change, the obvious explanation is multiple events, either two losses or one gain and one loss. Certainly if single events are so common, we should expect the occasional double event purely by chance. Not difficult at all.
This is a simple count of the data. Do I need to repost the picture that shows the data?
Hi John,
Since this argument started your case has rested on these 2 assertions.
As long as science restricts the analysis to methodological naturalism these assertions are fine.
If the game gets changed the data like in Sal’s flower and Winston’s dependency diagram will become a very big challenge for you and force a lot more rigor in your analysis. If indeed a mind is involved in the pattern a approach like @Winston’s is very interesting.
Competing theories that force more rigor are healthy for making progress. The data we are looking at is certainly not random. A known non random process like reproduction (common descent) is certainly going to explain part of the data.
How does introducing the supernatural change anything? Please be specific. What supernatural process would explain the nested hierarchy in the data?
How? What sort of rigor? Bill, this is more word salad whose purpose seems to be to make you feel good about ignoring the data, as you ignore everything I tell you.
Only half of all human genes are considered to have orthologoues shared with the depicted vertebrates. But hey, there’s four species in that figure so it’s rather silly to make any grandiose inferences about the extend to which orthologoues are conserved among all vertebrates. As you have had explained to you at least ten times by now, including more taxa will show many more genes are actually conserved, with different individual species showing selective losses. Many of the genes absent from Zebrafish compared to humans, for example, are likely present in many other fish. And that’s without considering mere sequence divergence as a factor affecting orthology annotation.
Another point is the term orthologue itself, as opposed to homologue. Orthologous implies the function is conserved (which without direct biochemical experiments are inferred by some algorithm based on sequence-similarity), but the species could well have homologues.
Why do you need to have all this stuff pointed out to you again and again? We explained all this stuff back when Sal first brought up that figure in that 5000-post thread.
I have to disagree. It implies only direct descent from an ancestral gene in a particular location rather than a duplicate in another location. A species without that copy could well have a paralog, and the paralog could conceivably even have the same function. And in the case of polyploidy, both copies could be homologs. If, during diploidization, a different copy were lost in two different lineages, they would still be orthologs. I’ll admit that the terminology in such cases becomes confusing.
The alternative explanation of the data is that animals were seeded by God on the earth fully formed. This fits the pattern in Sal’s flower as we start with unique gene sets and de novo genes observed.
The nested hierarchy does not explain these specific changes as you have previously stated. It only explains the pattern of change.
I am not ignoring the data.
As I said under the restriction of methodological naturalism the data is properly analyzed.
Okay. Then annoyingly this is another case where definitions can be a bit different depending on the source. But in any case, as you say, there very well might be paralogoues not counted by the algorithm.
How does that explain the nested hierarchy? Why should organisms being poofed into existence produce an objective nested hierarchy?
That’s because the nested hierarcy is a pattern to be explained, not itself an explanation. Like observing that temperatures increase and decrease over multiple consecutive 24 hr periods.
That up and down curve is a pattern for which we try to provide an explanation. The up and down curve is not itself an explanation. We explain the curve (the pattern in the data) with the night-day cycle. The sun sets, the backside of the Earth loses heat radiation to space, and then temperatures drop, then the sun rises as the Earth rotates, and heat enters the atmosphere from the sun, raising temperatures again. That’s how patterns in data are explained. The curve in our data is not explaining anything(though we can use it to generate predictions for future patterns in temperatures).
Gene gain and loss occurring on a branching tree (common descent) explains why there is a nested hierarchy(the pattern) in the conservation of genes among species. Gene gain and loss is an observed reality, organisms being wished into existence is not.
That’s not an explanation. Why should the data for these “seeded” organisms fit into a nested hierarchy? And a nested hierarchy is exactly the pattern in Sal’s flower.
Exactly. It’s the pattern of change that you need to explain. The causes of new genes and gene losses, though well enough known, are not relevant to the question.
What constitutes a violation of the “nested hierarchy”? Why would a system that is formed by seeded animals that reproduce not form a "nested hierarchy’?
The “nested hierarchy” is a label of the pattern. Were trying to explain a pattern or group of patterns that are not an exact fit to any labeled model.
The alternative question is why the gene sets fit into a dependency graph. You can claim they don’t perfectly as any model is not perfect. Where do we go from here?
A violation would be a pattern for a gene that doesn’t fit the tree, i.e. that requires more than one change. Homoplasy is the technical term. Of course homoplasy does exist, but in the figure we’re discussing it’s a negligible fraction of genes, easily attributed to chance loss of the same gene twice.
It could of course do anything whatsoever at the whim of the creator. But the real question is “Why should it?” And that’s what you have to deal with. Creation, per se, has no expectations whatsoever. Common descent does, and when those expectations are seen in the data, that’s evidence in favor of common descent.
Ah, but we expect a certain amount of homoplasy. Data are almost never perfect. What hypothesis best fits?
That’s obvious. The dependency graph was constructed precisely to fit the gene sets. A dependency graph has no expectations and can fit any data, including completely random data.
Where we go is where you have never gone: an explanation for nested hierarchy in the data other than common descent. If experience is any guide at all, I don’t expect any progress.
