Do you believe this process of descent from a common ancestor is solely from isolated populations and reproduction?
Hardly. My use of Australia is not as a model, but as convincing evidence that millions of years and isolation can demonstrate many a prediction of Evolution … without excluding the possibility that God was involved.
Australia does show YEC as not credible, and Old Earth Special Creation as a most strange “beast” indeed.
1 Like
Hi Bill,
According to ornithologists, the paleognaths are about 100M years old, and their dispersion is readily explainable by vicariance as Gondwana split apart. The problem is that this easy explanation is not consistent with genomic evidence. It was the genomic evidence, not the “youth” of the paleognaths, which compelled Hashman’s team to favor three loss of flight events within the ratites. I agree with Harshman’s decision to trust the genomic evidence. Not that my vote counts for much.
Absolutely not! Here are two very strong reasons to ascribe common ancestry to the ratites:
(1) The shared 8 base pair deletion in ALDOB (see Figure 3)
(2) The machine learning hierarchical clustering analysis, which yields a cladogram in which ratites have common ancestry.
I would like to take a moment to point out something very important: every day of the week, courts all across the globe uses the analysis of genetic similarities and differences to make inferences about common ancestry. That’s how paternity lawsuits get settled. That’s how distant nephews of great-grandsons lay successful claim to a share of an intestate bilionaire’s estate. It’s how executors of estates refute the false claims of purported distant nephews of great-grandsons. It’s how we know that Thomas Jefferson was the father of the children of a Jefferson slave, Sally Heming.[1] You measure the genomic differences and similarities, apply the mutation/inheritance model, and infer (or disprove) common ancestry.
My sister-in-law’s 23andme analysis turned up a big surprise: she (and my wife, we all presume) had about 1% Middle Eastern ancestry and 8% Berber. The whole family was initially stunned; my mother-in-law’s family had claimed Castilian roots, but not Middle Eastern and North African! Then we remembered: the Arabs conquered Berber North Africa. Then the Moors conquered most of Spain (Andalusia) and ruled it for centuries. That’s how significant amounts of Middle Eastern and Berber ancestry turned up in the genetic analysis.
Harshman’s analysis of paleognaths, volant tinamous, neognaths and crocodilians is based on the same principles and techniques that 23andme uses for human genetic analysis.
@Swamidass and @glipsnort - These last 3 paragraphs deal with your field of expertise. Please feel free to correct or clarify as needed.
Harshman offers 2 possible explanations:
(1) Vicariance: “Africa was the first piece of Gondwana to separate, and the African ostriches are the first clade to diverge.”
(2) Incomplete lineage sorting.
Given the paucity of fossil evidence related to the ratites, it is not possible to make a determination with 100% confidence. Harshman seems to favor ILS, however.
I would welcome any feedback from real biologists if I have misunderstood anything or explained anything poorly. I’m pretty good with statistics and building ML clustering models, but I could be a little fuzzy on their application to evolutionary biology.
Good questions, Bill. Warm Advent wishes!
Chris Falter
[1] Some of Jefferson’s modern-day descendants point out that the genomic analysis would also be consistent with Thomas’ brother being the father of Heming’s kids. However, the initial case for Thomas Jefferson’s role was based on the fact that each of Sally’s children were born about 9 months after one of Thomas’ visits to the estate.
3 Likes
Hi @Billcole -
A recent article in the Science section of the NY Times provides further evidence of the power of machine learning genomic clustering analysis. Munshi-South, et al., trying to understand the origins of NYC’s rat population, analyzed DNA samples from 314 rats from 31 cities. Read about their conclusions here. (Browse in privacy/incognito mode if you have trouble accessing the article.)
Best,
Chris Falter
1 Like
@gbrooks9 you were replying to a comment about natural selection as though the commenter were talking about mutation. Yes mutation is unguided and so can do damage to functional systems. But natural selection counteracts that by discarding the changes that are harmful and promoting those changes that are useful.
This is what is meant by “Natural selection isn’t random.”
Natural selection favours the “noise” that increases fitness and filters out the “noise” that decreases fitness.
1 Like
I love it when conversations here begin to revolve around actual scientific findings.
1 Like
Actually, all I was pointing out is that humans frequently refer to something as “random” that is 100% lawful…
or in reference to things that look random to them, but which are not random to God.
I recommend being very careful trying to prove something just because you want to call something random or non-random.
Here is a source that starts the breakup 80 million years before the paleognaths. Can you help reconcile this? Per Livescience
quote The shared 8 base pair deletion in ALDOB (see Figure 3)
(2) The machine learning hierarchical clustering analysis, which yields a cladogram in which ratites have common ancestry.[/quote]
While I agree that the 8 base pair deletion is evidence that supports common ancestry (except the ostrich), the cladogram simply sorts the data and arranges a tree and assumes ancestry based on closest DNA match.
At this point, unless we have the rest of the DNA sorted, I don’t think we can estimate the feasibility of common ancestry based on isolated populations and genetic recombination.
I think ILS means incomplete data. How do you conclude common ancestry on incomplete data unless it is an a priori assumption?
Hi Bill,
Interesting discussion, I’m enjoying this.
Well, the molecular clock analysis points to a common ancestor in the early Cretaceous, which still (barely) fits the hypothesis. Africa didn’t break off until the early Cretaceous. You can see Africa still connected at the end of the Jurassic in this graphic from the USGS:
The competing hypothesis is that multiple ratite phyla independently lost flight starting at roughly the K-Pg boundary. This is plausible timing, as the disappearance of large dino predators would allow the disadvantages of flightlessness to diminish and the advantages to be accentuated. The hypothesis is also supported by Harshman’s analysis, and by the ratite fossil record, thin as it is.
Agreed.
This is the exact technique used by 23andme for human ancestry. It’s very effective and accurate.
It’s true that Harshman didn’t use the entire genome. Their sample size was quite large at 14 taxon and 23,902 base pairs, though. Their methodology section shows that their Bayesian analysis considered the null hypothesis appropriately.
Have you done a statistical power analysis that would show their sample to be too small? I would be interested in seeing your numbers. In fact, you need to back up your assertion with numbers, Bill. Whether or not the sample size was too small is something that can be determined quantitatively. It is quite unscientific to assert insufficient statistical power without performing a quantitative analysis.
In fact, as Dennis Venema explains, ILS is an expected phenomenon and it shows up in genomic data in the expected frequency.
Expressing an opinion and adding $2 will get you a cup of Starbucks, Bill. But it isn’t science. If you think there’s a better way to model the genomic data, Bill, you should go for it. Harshman tells you how to find the data:
Data deposition: New DNA sequences are deposited in GenBank (accession nos. EU805776–EU805796, and EU822937). Alignments and trees have been deposited in TreeBase (study accession no. S2138).
Download RStudio if you haven’t already, download the data, and tell us what you find. I’d be interested to see what you come up with.
Advent blessings,
Chris Falter
1 Like
Beat me to it! Yep, it would be strange if we didn’t observe ILS.
1 Like
You and I and every scientist is always working with incomplete data, Bill. For example, we have incomplete data on the theory of gravity because we have not observed every gravitational attraction since the beginning of the universe. But our sample size is large enough to give us confidence about the theory.
In a universe of incomplete data, do we just throw up our hands and believe whatever tweets we like? No, we use statistical power calculations to determine the validity of scientific analysis.
Advent blessings,
Chris
2 Likes
Hi Chris
Thanks for the education on plate tectonics and ILS. Can you make an argument that ILS does not assume common ancestry?
In the case of humans we can assume ancestry. Can you make an argument that this technique establishes common ancestry when it is not assumed?
I am not sure that our current understanding of the genome would allow us to do this accurately. If you think I am wrong, I would be interested in your argument. At this point some scientists believe that 10% of the human genome is functional while others believe 80% is functional.
Chris, with all due respect this statement is non sense. Sharing ideas freely is a very important part of science prior to coming up with a testable hypothesis. Do you think there is a testable hypothesis that establishes paleognath common descent by reproduction?
There is a testable hypothesis that would demonstrate NON-common descent…
… and that’s by showing that paleognath “A” does NOT share the same genetic markers in stable areas of their genome that paleognath “B” has.
Is this a surprise to you?
1 Like
In the case of gravity we have a tested mathematical model. You are comparing apples with oranges.
We are comparing apples and oranges if you know of countervailing examples where Genetics completely indicate common descent and we know for a fact that a child is born by some other mother …
Or if genetics indicate completely different parentage, and we Know for a fact that the child does share a common mother.
You are tilting at windmills, Bill.
1 Like
We know we have common descent among species. Among different species is where the challenge is. What mathematical model can reconcile how a man and chimp evolved from a common ancestor?
Take a look at this well-known case:
Is there anything about this that you think is “hinky”? Why would God arrange chromosomes like this ? To deliberately fool humans into thinking Chimps and Humans had a common ancestor ?
Your belief that Species are somehow special islands incapable of drifting into new forms is a rather forlorn position. Australia is full of marsupial mammals who have drifted into roles that are notoriously filled by placental mammals in the rest of the world.
Tigers and Lions are known to cross-breed? If these creatures are really just ONE type … what do you think is going on in the animal world? Are whales just another type of Hippo?
More tilting at windmills…
1 Like
Hi George
I honestly don’t know how much we can attribute to common descent and the merger of the chromosome is evidence but the reconciliation of 22 million fixed mutations and possible 30 to 40k splicing variations is equally a challenge on the other side of the argument.
How all this and how de novo proteins are formed through genetic recombination is not well understood and is very possibly wrong.
I would challenge that you are making pre mature conclusions based on insufficient evidence.
This however was not my original argument that you challenged. I claimed that a mathematical model for the chimp to man convergence from a common ancestor was unlikely and therefor my apples to oranges point against the theory of gravity was valid. Do you now concede this point?
I do not concede that point. Probabilities are not my strong point… however !..
I’m confident that probabilities could be calculated for the likelihood that 2 random Pre-Chimp chromosomes (for the purposes of this exercise, we would use today’s Chimp configuration) could fuse, vs. 2 specific Chimp chromosomes could fuse to create the human configuration… sustantially confirming the realistically low probability that chimp chromosomes and the human configuration are accidentally similar.
And when you combine this with the oddity that only a small number of primates have a functioning gene that allows the synthesis of Vitamin C (like the great majority of Mammals!) … we get an additional set of probabilities signaling another level of common ancestry. This is especially interesting because we have all the other genetic code for making the Vitamin C … that we, like most of the other primates, cannot make.
Isn’t it miraculous that God has provided scholars with some of the most interesting “coincidences” specifically for Primates … the one group of animals most interesting and relevant to the study and investigation of Evolutionary developments. I think that is a truly Divine Coincidence!!!
1 Like
Hi Bill,
How’s this: ILS has been observed within populations known to have common ancestry, such as humans. Does that help?
By the way, what did you think of Dennis’ articles?
False. The technique does not assume common ancestry.
In the case of Jefferson and Hemings, the Jefferson clan and the Hemings clan made differing assumptions about common ancestry. The technique adjudicated their differing assumptions.
Your statement is quite orthogonal to the question of statistical sampling and statistical power. Have you done any university-level work in statistics?
You already accepted paleognath common descent based on the 8-bp indel in ALDOB, I thought.
So what is the problem here? Are you contradicting what you wrote earlier? I’m not following your thinking, I’m afraid.[quote=“Billcole, post:54, topic:16409”]
In the case of gravity we have a tested mathematical model. You are comparing apples with oranges.
[/quote]
I already explained how the genomic mathematical model works very powerfully with a known case (human lineage). Biologists apply this tested model to other creatures who reproduce in a similar manner.
Blessings,
Chris
1 Like