It’d be better you let Jeanson defend his work himself, because you don’t seem to be doing well defending it. Its too obvious you are out of your depth here. If you still want to go on its fine, but just know its getting embarassing for you.
I’m going to be very blunt about this.
You are wrong.
Jeanson is wrong.
You can not infer the sort of population dynamics Jeanson claims from these trees. At all.
There is so much wrong here. For starters, in a large exponentially growing population most of the novel haplotypes will be present In only one copy in the population and will likely go undetected.
In fact any meaningful analysis based on fundamental population genetics show just the opposite of what Jeanson is claiming. It’s right there in the papers he cites and he totally ignores it just as you are doing.
Let’s be honest. Please. Have you ever thought that you don’t know enough about this to have an informed opinion about it?
There are many short branches at the tips because of a BOTTLENECK not population growth. Look at how those branches line up with the Bayesian skyline plot in figure 2. They correspond to a small population. Not a large population. It’s exactly what I explained in my video with Dan.
Again, it’s about evidence, not words.
Instead of grasping at textual straws, you’d be more convincing if you examined the data.
In fairness, I think she has at times been quite open about this. The way it seems to work for her is she realizes she really does not understand much of this stuff at all. so she can’t really understand the arguments being made against Jeanson’s claims. Therefore she is not in a position to say that he is wrong.
Sure, I need to take the big L earlier because I was very frustrated, too hasty in making assumptions because I hadn’t read Karmin or Jeanson’s paper in a while, I’m not trained in this subject and learning poorly as I go, and just jumped at the fact that the general shape of Karmin’s tree looks like a coalescent tree of constant size that @Herman_Mays was talking about. Not fun to make a big mistake especially now when I know better.
I did take the time to read over Karmin’s paper and the supplemental information and Jeanson’s paper again because it’s been a while.
Ok, so I agreed with you that Jeanson’s method doesn’t work in coalescent trees, I will grant you that it doesn’t work to analyze Karmin’s tree, I grant you that it doesn’t work in trees that have rare mutations because novel haplotypes will go undetected (which is the 2nd question I had).
ALL that I was trying to say is that you have not ruled out that Jeanson’s method can work for the distance based tree he created with Karmin’s data with a mutation rate as high as he claims. That’s it. That’s all.
No, again I’m saying that I can grant Jeanson used the wrong method for everything else, but that he, as far as I can understand, used a method that’s okay for his tree as long as the mutation rate is high and fairly consistent. And he’s showing it’s working, so why is that? He should not be able to continue to do that if the mutation rate is far, far lower.
OK. So Jeanson’s methods don’t work with coalescent trees. So the question to you is why did he use these trees? Why did he take the Karmin et al tree and apply his method to it if he knew it didn’t work? Jeanson’s neighbor joining tree from the VCF files he borrowed from Karmin et al are basically the same tree with a similar topology. He applied his method to BOTH TREES. He made NO DISTINCTION BETWEEN THE TWO. Counting branches as a proxy for population size doesn’t work on ANY GENETIC TREE.
Jeanson has absolutely no idea as to what he’s doing.
Valerie. There’s obviously nothing I can say to you to convince you that you are being duped. That’s an exceedingly difficult thing to do in this situation no matter how much time I take to dissect Jeanson’s work for you because Jeanson’s conclusion is what you want to hear and mine isn’t. Jeanson’s beliefs about human history conform to your own and affirm your religious convictions, mine don’t. I’m here to tell you however, as resistant you are to this, that as someone with professional training and responsibility in this field that nothing Jeanson is doing is justified scientifically. None of it.
Okay. You do realize it’s not working, right? Nothing Jeanson says or does is valid, from the way he “calculated” mutation rates, to deriving population growth from the number of branches on a tree, to the times when various haplotypes supposed appeared in various geographic locations. It’s all completely, demonstrably wrong. Why can’t you acknowledge that?
And think about this Valerie.
The coalescent approaches, the sorts of methodology used throughout population genetics for decades, tell a story completely at odds with Jeanson’s approach, a method never before used in population genetics.
A coalescent approach is one grounded in basic population genetics and that methodology shows that short branches are a consequence of a small population, not a large one. I’ve explained this and it’s right there in the Karmin et al paper in figure 2 showing that the short branches of the Y chromosome tree correspond to a population bottleneck NOT an exponential expansion.
So which approach makes more sense as the right one? Both can’t be right. So which is it?
A set of methodology grounded in empirical findings and basic population genetic theory used across thousands of peer reviewed research articles for several decades.
An entirely unique approach invented by someone with no training in population genetics published not in the scientific peer reviewed literature but online on the website of a religious ministry that just conveniently tells him exactly what he is contractually obligated to believe.
For subjects that are simply too complicated to grasp in course of a social media thread or Google search it’s sometimes good for the general public to have some pretty rudimentary heuristic tests of credibility, a smell test if you will.
In the way I have framed this problem above ask yourself honestly if what Jeanson is doing passes the smell test. Try and separate this issue from what you would prefer to believe.
Ok, but can we put coalescent trees aside for a moment?
Let’s take my family for instance. My dad has three sons, the oldest has four sons, the second has 3 sons, and the 3rd has four sons. (We are lacking in granddaughters on my side of the family. ) I drew a family tree from this data.
Then let’s say I sequenced the Y chromosome of these particular nephews and their dads and I’m obviously aware of their genealogical relationships already. And I noticed there were 3 differences per generation.
Then I kept just the sequence data of my nephews only and I created a distance based tree. Assuming there are no back mutations, is it or is it not going to look exactly the same as the family tree I drew earlier? Is it or is it not going to show that from the 1st generation to the 3rd generation the male population increases 11x?
Not really, no, we can not set aside coalescent theory, because coalescent theory lies at the very core of any modern attempt to infer population dynamics from genetic data so you can’t just say it doesn’t exist. You really can not infer population dynamics without considering these approaches.
Where is Jeanson claiming these sequences come from any one family or single generation? They in fact don’t.
You are focusing on one path in one family. The reality is these sequences are taken from a more-or-less random draw from across multiple populations from around the world, not from any single family. Sure your family has many males in those generations but each one of those individuals has other ancestors too and those parts of this family pedigree may be entirely different. Another family in the same population may only have a single son, or no sons at all.
Let’s instead say you take Y-chromosome sequences from a stable population. Some sequences come from large families with many sons. Some sequences come from small families with few or no sons and they are sampled across multiple generations. If the population has only one son each generation per male parent on average then it is neither growing nor shrinking with regard to the Y-chromosome. Do you really think that if you sample from across males in this population they will all be the same sequence? With no branching tree to be found?
Population growth in Japan, for example, has been stable or declining since at least 2010 yet you still could collect Y-chromosome sequences from Japanese men and those sequences would in fact vary and you could display that variation on a bifurcating tree.
Jeanson uses basically the same example from his own family (ignoring the fact that he has daughters) and says if he and his wife have four kids then the population in the next generation is four times larger. No, it is not. Another family may have no kids or one kid. On average in a stable population, each individual is being replaced, no more no less. That doesn’t mean genetic variation disappears.
Plus human populations are subdivided. Some populations are growing, some are shrinking, etc. and you don’t have an equal chance of contributing genes to any population.
Genetic trees are basically gene genealogies, not pedigrees. Gene genealogies can be found in any population whether they are stable, or growing, or shrinking and coalescent theory tells us what we should expect those geneaologies to look like. Jeanson understands none of this.
You’re not seeing/answering the basic point I was trying to make. If I drew a patrilineal family tree of those three generations, then I created a distance-based tree from the grandsons’ data in that scenario, AFAIK they would look/be the same.
If that’s true of my tree, that’s true of the entire y-chromosome tree. It’s just one big patrilineal family tree, a male pedigree tree.
If that’s true, then Jeanson’s genealogical pedigree approach works just fine:
- the world population has grown exponentially; this tree will have long ancient branches and many short ones at the tips
- the tree will indicate minimum male population size
- the tree won’t just bifurcate, it would have a branch or branches for each son that has a surviving male line
- you could overlay the tree on any historically and biologically accurate male genealogy
- finally, if you have a random sample of the population from that tree, and create a growth curve from the internal branching, it should match the population growth curve, if your sample is large enough. (In the book Jeanson was matching the curves over several millennia and is not dealing with a stable population at the tips of the tree)
Sorry to belabor this point again; I just think if I was in your shoes, I would dismiss what Jeanson was doing. If you can put yourself in mine, you can probably see that you need to address the shape of his tree to rule out his approach, as I’m already aware his approach is novel.
Is it consistent? Do you define that quantitatively, or just with words, to give yourself wiggle room?
He’s conning you into thinking it’s working. That’s different from showing. Showing, in science, typically involves testing hypotheses. Creationists only pretend to do that.
I’d like to see the math you used to convince yourself that it is working. From you, not from pasting text.
What’s the probability that those differences would be evenly distributed?
Jeanson is conning you.
A pedigree is a reticulate network not a bifurcated tree and yes I do see the point you are trying to make but unfortunately it doesn’t work. You are zooming in on one line and one line only and forgetting that this is a population and not every father is having as many sons. If the population were stable you could still build a bifurcating tree from the sequences you collect from the population. You and Jeanson would wrongly conclude the population is growing merely by observing a tree.
His approach is simply wrong. That’s it. I’m not merely dismissing anything. Whether the population is stable or growing or shrinking you would get a tree from a comparison of sequences you pull at random from the population. That’s the point. Bifurcating trees do not only exist for growing populations. If you want the “shape of the tree” to say something about population dynamics then you have to think about this from a coalescent approach and in that case you get the opposite of what Jeanson claims. And I’ve explained why. The short branches come from small populations. The long branches from large populations. It’s all the exact same data. You can’t take a likelihood-based coalescent approach and say the population is doing one thing and a neighbor-joining tree FROM THE EXACT SAME DATA and say it’s doing another. Those two trees are in many respects that are important here are similar anyway and remember, JEANSON USED TREES FROM BOTH APPROACHES. He doesn’t know how to infer population dynamics from ANY tree. That’s the problem.
Ughhhhh…I’m quite frustrated with myself, as I thought my confusion earlier was the result of Jeanson using different data. You had said early in the conversation Jeanson’s tree was a distance-based tree, and I was thinking he used the same data. If his tree was based on multiple mutations per generation, I thought it couldn’t look any different than having short branches at the tips. But so did Karmin’s - so when I realized that was a coalescent tree, not a distance-based tree, I thought I was wrong about the data because the trees looked the same. I was just confused twice.
OK, but I was just trying to chose the right words to make sure we’re talking about the same thing. I don’t know how else to describe a male-line only family tree that just shows my dad, his sons, and their sons and no one else in the family. I don’t know how biologists use these terms. I couldn’t tell from your replies if we really were referring to the same thing especially because you keep using the word “bifurcating.”
I was checking out what the tree looked like in a stable population earlier this week using paper and pen - you definitely get branching. Later I realized my branching wasn’t realistic because I had no childless males - one each had 2 daughters or two sons, and two had one of each. You can easily tell drift happens really quickly.
But as far as the branching in a stable population goes, that’s why Jeanson is saying the tree shows minimum population size. Obviously that has to be true even in a stable population.
Today after I realized how I had been confused again about how the trees look, I knew I really needed to understand your point about the shape of coalescent trees. The language you used in the critique video confused me since you were referring to genes. I found this great little video.
I realized for a simulation in this scenario I can just think of the dots as men, not genes. All of the simulations in the video showed a stable population. But I thought - that’s not realistic. Exponentially expanding populations are the norm for human history. But what we’re really contending about is whether Jeanson’s right about what the tree looks like in declining population. So I got out my paper and pen again and first tried out a slowly declining population. I put my phone over each previous line as I drew the second generation so it would be more random. And then I drew the y-chromosome tree as it would be according to Jeanson. See the first picture.
Then I thought…that’s not realistic either. Let me try a bottleneck - war or disease would cause a rapid population decline, but then it would quickly grow again. See second picture. (Please ignore my mistakes; I had a terrible eraser that didn’t work. I’d like to make it look like I can spell the word “branches” at least, but I’m too tired to redraw it )
So I’ll make the obvious point - the bottleneck turned out like his inference about the shape of the tree. And the number of branches followed population size once the population began to grow again.
Somewhere in the middle of that, I realized that your example in the video with the distinction between large and small populations and their probabilities and what the tree then looks like, assumes completely random mating. But we know that is not the norm: Humans tend to act like they are in small populations, even if the entire population is large. A college friend of mine shared a joke on Facebook a while back how she and her mom were looking into their ancestry and found out they were 3rd cousins once removed. (Her parents were 3rd cousins and didn’t know it). So maybe all trees should have short branches at the tips if the mutation rate is high…but maybe coalescent theories take all of that into consideration already.
Valerie you have to do more here than sketch things out by paper and pen.
A bottlenecked population will have one or a few long branches and lots of small branches at the tips.
The shorter branches arise DURING THE BOTTLENECK WHEN THE POPULATION IS SMALL AND NOT WHEN IT IS GROWING AFTER THE BOTTLENECK.
When you connect two genes on your trees how do you decide if they happen to join together in the previous generation? I’m trying to get you to understand this, for what it’s worth…