With deliberate nastiness like this, it’s no wonder this forum is such a one-sided echo chamber. Why would anyone bother?
You call that nastiness? You have led a sheltered life. That was a summary opinion of the quality of your arguments. Nothing against you as a person.
5 Likes
My call here is that John is expressing an opinion, and not a particularly “harsh” opinion as these things go. There might be a lack of understanding about the intent, but that likely started before these comments.
2 Likes
I have seen deliberate nastiness on PS. This is not it. Not even close.
3 Likes
How is this “deliberate nastiness”? It would seem to be a reasonable perspective given:
-
your and Dr Hancock’s relative expertise in population genetics; and
-
the respective depth to which GE and mainstream population genetics have developed their theories.
This could better be attributed to the mental wear and tear of defending the indefensible to a well-educated and skeptical audience on an ongoing basis.
Only the most masochistic of apologists attempt the task.
Most apologists quickly return to apologetics’ forte – preaching to the choir – still “a one-sided echo chamber” – but a more salubrious one.
3 Likes
Please let’s drop this point of discussion. If @UncensoredPilgrims needs to comment on it he can PM me.
1 Like
You win just for using “salubrious.”
3 Likes
Anyone else watched this debate yet? I think my prediction was spot-on.
It seemed to me that Paul basically ignored the real world evidence that real populations dont decay and spent the debate arguing from selective quotation and how he could produce the effects of GE with a custom fixed (they arent fixed in reality) DFE in a SLiM simulation. Though I agree with Zach it is commendable that Paul went to the length of learning SLiM and trying to actually model GE using non-fraudelent software.
There were some technical points about the relationship between effective and census populationsize between Zach and Paul I didn’t understand near the end, which there wasn’t time to elaborate on. Perhaps @talkpopgen can clarify here if he is around?
2 Likes
Census population is the total population. Effective population would just be the ones that reproduce, which is smaller, sometimes very much so. And there may be separate male and female effective population sizes.
Yeah but there was a dispute in the debate near the end, about whether it was appropriate to use one or the other regarding a specific question, and then more disagreement followed from that. Which there wasn’t time to further clarify as they had run out of time.
This specific dispute appears to start at 2:07:52 in this video:
Zach clearly becomes quite exasperated with Pauls response to what he says there, and there’s some back and forth, but I’d be interested to see some more clarification on it especially following Paul’s last response on that subject.
Effective population size is perhaps the most misunderstood concept in all of population genetics. In a nutshell, it is a measure of the rate of genetic drift. Slightly more technically, it’s the measure of drift relative to expectations under a Wright-Fisher model of constant size N. In such a model, drift is caused by random binomial sampling and has a per-generational variance of p(1-p)/2N.
At equilibrium, expected diversity is 4Nu, where u is the mutation rate. When the Wright-Fisher assumptions of constant size, random mating, etc. are violated, genetic diversity might differ from this equilibrium expectation. For example, if the population size is not constant but fluctuates, then Ne is the harmonic mean of census N.
Human Ne is often cited as 10-25,000 based on this equilibrium definition, which is commonly called the “inbreeding Ne” or “long-term Ne”. This can be instructive about how selection has acted in the past, and how much diversity is present to be acted upon in the present. My argument in the debate, and in the video that Paul is citing, is based on the variance Ne, not inbreeding Ne, which is defined by 1(1-p)/2N, or, in words, the expected amount that an allele will fluctuate in one generation by chance. The reason this definition is more relevant is because we’re predicting selection into the future, not in the past, which must necessarily be shaped by current levels of fluctuating gene frequencies. These fluctuations should occur at a rate closer to the census N than the historic one. Consider how long a new mutation would take to reach fixation in the human population, or the rate at which it would change from one generation to the next - obviously, this will be much closer to N = 7 billion than N = 10,000.
Kondrashov used the long-term Ne because he was assuming an equilibrium population size in the past. But Paul rejects such a thing exists; however, we clearly exist in the present, so if we want to project the future we should use variance Ne, not equilibrium Ne, to determine the strength of selection and effective neutrality.
That’s the basic context I didn’t have time to explain in the debate. Hope it helps!
5 Likes
Incorrect. I do not deny the past exists, and I clearly do accept that population sizes have been much smaller for the vast majority of our history as a species. Long-term Ne is determined, among other things, by the harmonic mean of the population size over time. NOT “equilibrium”.
Not only that, but most species on the planet have not exploded in size over the past centuries as humans have (due to the industrial revolution). GE is a problem for them, too.
This claim about “predicting the future” is a red herring. We’re talking about whether evolution is a viable explanation for origins, not just trying to predict the future. You’re mischaracterizing the literature in your own field. Nobody thinks the human Ne is anywhere near 7 billion.
Well you’re discussing both. When you create a model in SLiM and try to simulate what will happen to the human (or any other species) population going forward (will we go extinct due to GE?), you are in fact trying to predict the future. So it clearly isn’t a red herring there.
So there are both questions about how we got to the present from some state in the past, and what will happen in the future going on from now. You appear to be claiming that Zach is wrong about what value for Ne will be most appropriate to use for a simulation trying to predict whether the current human population will go extinct due to GE in the future. Right?
2 Likes
That’s a dodge. The point of the simulation is to show that the dynamics of realistic genetics does not comport with evolutionary claims about origins and the past.
Again, it’s more about the past than the future (why have we not died 100 times over??), but, he is wrong on two different counts:
-
His claim that Kondrashov’s paradox is based upon current or recent Ne is wrong (it’s based on long-term)
-
His claim that the current Ne is anywhere near 7 billion is also wrong. That completely ignores the issue of population substructure and of Hill-Robertson interference, as Dr. Lynch pointed out. And on top of all that, not all living humans are even of an age to reproduce.
No it’s a perfectly reasonable explanation for why both questions matter. What a strange claim to say it’s a dodge.
Are you saying you don’t predict extinction due to GE in the future for the human population then?
I am quite perplexed about this retreat from discussing predictions of future extinction due to GE. It looks like you’re the one dodging to be quite frank.
1 Like
It’s not a “retreat”. Hancock attempted to reframe GE as if it were only about the present and the future (in order to justify changing which version of Ne he would look at). That is manifestly wrong, and as I showed in the debate, you can read it straight from the source, Kondrashov’s own paper, to see that. Of course I do believe all LMEs are on the path to eventual extinction, unless God intervenes first.
Can you point out a timestamp in the debate whether this occurred? I watched it and saw no such attempt. I’m sure I’m just biased towards his side though, but will need your help for that reason. Thanks.
I’m not sure whether he said it in the debate, because during the debate he ignored pretty much everything he ever said previously about genetic entropy. A curious tactic, to say the very least. But you don’t need to look at the debate, you can read what he wrote just now, here in this forum.
This too, contains subtle misdirection. “Equilibrium size” never had anything to do with it. Long-term Ne is determined by many factors, but largely it comes from the harmonic mean of the population size over time, not some kind of equilibrium size.
Well surely if you’re modeling GE going forward from the present, it really does matter which Ne we use right? So you’re saying if we attempt to model GE going forward from the present, we should still use Ne of ~10K, as opposed to something closer to the census population size. And when a population geneticist tells you we shouldn’t use the historical value resulting from a bottleneck in the past, you reject that (with an appeal to your comprehension of Kondrashov).
I wonder though, if we got other population geneticists to weigh in here. Would they agree with you? What would it take to persuade you you’re the one who’s wrong?
I didn’t start at the present, I started at the beginning of humanity.
If I have learned anything about this forum in my time with it, I have learned that nobody here is ever going to agree with me, especially when I am right.
Considering that you’re currently ignoring all the evidence I have presented, I find it ironic you would be asking me that question.