My mistake. I incorrectly thought that we had collapsed the subject into mere sequence analysis.
I think the discussion started out there. DNA sequence analysis as the strongest evidence for common descent. And from there I think I took it into phylogenetic trees thinking that’s how DNA sequence analysis is actually used as evidence for common descent. I may have failed to make an important distinction.
1 Like
Phylogenetic trees are not nearly the only or main way that DNA supports common descent.
2 Likes
You speak, perhaps, of bootstrapping or of the number that MrBayes assigns to a node; neither of them is a p value, really. Neither is rejection of a null hypothesis.
There are different sorts of sequence analysis, and it’s the phylogenetic analysis of sequences that provides the best evidence of common descent, as you say.
Such as?
-
The first way I was exposed to neutral theory was by Ka/Ks ratios. When I saw the logic of it, I was (1) stunned I’d never heard about it from ID, and (2) saw how it undermined entirely the argument against CD. They had just refused to engage with it. This is really profound evidence, even though you only need two species to compute it, not tree involved.
-
Likewise, the ratio of transitions to transversions in four lines of very strong evidence: inter-species divergence data vs. intra-species variation data (both number of rare variants and allelic spread) vs. in vitro/biochemical data.
-
In another distinct test of variation, the correlation across different genome loci between mutation rate, divergence, and variation (both number of rare variants and allelic spread).
-
Independent of this, also, is the test of phylogenomics for inferring function (contra @pnelson and @stcordova) versus sequence similarity.
One could argue #4 is phylogenetics, but I’m not sure it is only phylogeny. Sure, to compute allelic spread, we need phylogenies, but this also is just part of the story. Even leaving this out, there is sufficient data without even that line of data. Of course all these tests rely on DNA analyzed without a tree.
3 Likes
I will admit that I don’t clearly understand how that’s evidence for common descent. Unless you mean that the differences between species are the sort of thing we would expect if our models of purifying selection were true.
This does, however, demonstrate my point that everyone thinks the strongest evidence for evolution comes from their own specialty.
1 Like
Not really. I mean genetics very broadly. This was just my personal first taste of it. Others have different entrypoints. If you notice, I very rarely mention Ka/Ks ratios, because it is obscure even if it is convincing. I’m just saying there is more than merely phylogenetic trees. Also, Ka/Ks is not really part of my scientific work any ways.
But clearly you’re more into that than phylogenetics. Have you ever tried to discuss Ka/Ks ratios with a creationist to convince him of common descent?
I talk about phylogenies too @John_Harshman. You know, maybe there isn’t actually an argument here to win?
Of course there’s an argument. That’s the title question. It’s just that the answer isn’t important, since there’s enough evidence to go around in a dozen or more fields.
1 Like
Not if both agree that it is genetic evidence of one sort of another, which we do. We both agree on the answer. I’m just showing you how that answer extends beyond merely phylogenies.
What do you mean “merely”? Them’s fightin’ words.
2 Likes
Edit at as you please. “Only”? You are bruising for a fight, aren’t you? Is it uncomfortable to find yourself agreeing so much with a Christian in science? 
I refuse to use emojis to signal humor. The words ought to be enough.
2 Likes
Regarding Ka/Ks:
https://onlinelibrary.wiley.com/doi/abs/10.1002/bies.20592
While it has often been assumed that, in humans, synonymous mutations would have no effect on fitness, let alone cause disease, this position has been questioned over the last decade. There is now considerable evidence that such mutations can, for example, disrupt splicing and interfere with miRNA binding. Two recent publications suggest involvement of additional mechanisms: modification of protein abundance most probably mediated by alteration in mRNA stability and modification of protein structure and activity, probably mediated by induction of translational pausing. These case histories put a further nail into the coffin of the assumption that synonymous mutations must be neutral.
What is your point? Nothing new there.
We have know for a very long time that synonymous mutation need not be neutral. As evidence for common descent, they only need to be neutral more often than non-synonymous changes. That’s it. This is indisputably the rule.
Of course there are exceptions. Balancing selection at the HLA locus means that non-synonymous changes are more likely to be positive, which shifts the balance, just as Kimura predicted about 50 years ago. For every rule in biology, there is always exceptions. This exception, historically, was key to proving the rule. The fact that they could, from theory, predict where the exception would arise was very convincing.
So why would @stcordova make this objection? Probably for pedagogical reasons, so we can all understand why this is such a robust result. At least this is my most charitable guess.
I provided the link for completeness to see if the data was possibly relevant. I do not know for a fact whether it was relevant or not. I provided it to see if anyone found it meaningful and relevant. At the very least, I thought it was informative.
I think the following data point, may or may not be relevant, but since we are talking Ka/Ks. I’m just putting out the data point to see if it is relevant or could lead to another productive conversation in another thread if it is not relevant to the present discussion.
When I first met John Sanford, this study was of interest to him.
A low Ka/Ks ratio indicates strong selection; conversely, a high ratio, weak selection. Some genes have a ratio of 0, which means protein changes are not accepted. It is, in a sense, “perfect.”
For a pseudogene–a stretch of DNA sequence that resembles a gene but has no function–its Ka/Ks ratio is approximately 1.0, which means that synonymous and nonsynonymous mutations are accepted at the same rate since the gene is functionally irrelevant.
For a gene that is highly functional and important for the organism, its Ka/Ks ratio is typically low. For example, if a gene has a Ka/Ks ratio of 0.1, it means that it’s highly selective and is only accepting 10 percent of the nonsynonymous mutations.
Regardless of the rate of new mutations at a particular gene, scientists have always presumed the percentage of nonsynonymous mutations accepted during evolution remains constant.
“This theory has been the workhorse of molecular evolution,” Lahn said. “Thousands of scientific papers have been published based directly or indirectly on this notion.”
The new data show that if more mutations show up at a gene, that gene tends to accept a higher percentage of those mutations.
1 Like
I think some of the strongest evidence in favor of evolution are the shared errors in the genomes across species, the nested-hierarchical patterns of similarity and diversity, the progression from simple to complex creatures as well as radiation of forms.
2 Likes