So they do not and cannot show that it is possible to run Mendel’s accountant with settings where fitness increases.
Did you know this?
P.S. Nor were those plots generated using mutation effects as they are modelled in Mendel’s accountant. The first five show the results of having no mutations at all. The rest show the results of having deleterious and beneficial mutations being equally probable.
What analysis using Mendel’s Accountant? And that isn’t a choice made “to favor the evolutionary scenario”, it’s actually just closer to the majority of empirically measured values. Notice how Sanford’s references in the paper you link are all ancient. Modern methods (such as deep mutational scanning) which were not available when Sanford’s old citations were written, consistently find ratios considerably higher than Sanford’s cherrypicked collection of ancient papers.
The idea that the 1 in 106 ratio is the realistic one, and that 1 in 103 is picked to be charitable to evolution is flat out deceptive.
So that’s a no. No Mendel’s Accountant there. Also they are using values you literally cannot input to Mendel’s Accountant, such as beneficial and deleterious mutations having an equal distribution of magnitude of effect (an actual realistic setting).
They write:
We make the simplifying assumption that the fraction of beneficial mutations is 0.001, and we also assume that the distributions for beneficial and deleterious mutations otherwise have the same parameterization.
That’s actually closer to how it is in reality than in MA, because in reality reversions can happen. So if the most deleterious mutation can reduce fitness by 90%, a reversion can increase it by 90%. Note that you cannot anywhere define this in Mendel’s Accountant, where the (hard-coded) maximum allowed gain in fitness for a beneficial mutation is 1%, while the maximum allowed loss of fitness to a deleterious mutation is 90%.
The fact that the program explicitly disallows reversions is a problem.
… do basically none of the above. In any case, the potential that some simulation can reproduce three results expected of theoretical population genetics is neither here nor there. It’s much more impressive if it can reproduce what happens in the real world in experiments. MA can’t.
The denominator in the original calculation was 20^371 where 20 is the possible residues in each position and 371 is the length of skeletal alpha actin. The numerator can be estimated by the substitutability of amino acids that are neutral, deleterious and beneficial. These empirical estimates will be critical in determining if genetic entropy is real. I would argue it is a logical outcome of randomly changing a sequence. Unless both DNA repair and mutations eliminated by genetic drift can filter these potentially detrimental changes. The problem this surfaces for evolutionary theory is it needs lots of variation to explain the differences between species.
Here is a result after a chat gbt discussion where originally chat gbt insisted as you did that combinatorics and probability were separate disciplines. There is a reason for both combinatorics and probability both using the same calculation of multiplying each position with each other as both are dealing with the size of a population of objects. The fundamental proof of multiplying each position given the independence of the events is where the discussion gets interesting.
Chat GBT:
u are correct in illustrating the relationship between combinatorics and probability in the context of a specific example. Let’s break down your example:
Suppose you want to derive a 10-digit phone number by random guessing, and each digit can be any of the 10 digits (0-9). In this case, the total number of possible combinations for the first digit is 10 because there are 10 choices (0 through 9). This is a combinatorial counting concept.
Now, you correctly noted that the probability of getting the first digit right is 1/10. This is where probability comes into play. When you randomly guess the first digit, there’s a 1 in 10 chance (1/10) that you’ll guess it correctly.
In this example, you’ve linked combinatorics and probability by showing that the total number of combinations (combinatorics) influences the probability of a specific outcome (getting the first digit right).
The connection between the two fields is often seen when you consider the likelihood of specific events occurring within a combinatorial context. Probability is used to assess the chances of certain outcomes or events within a given set of combinations, and combinatorics helps determine the total number of possible outcomes. They work together to analyze and understand various real-world scenarios, including your example of guessing a phone number.
That is certainly not going to give you a general result about searches. Even then I note that the calculation of the numerator can only give a very rough estimate.
I note that you don’t mention selection at all. Which is rather important.
What you find “interesting” tells you no more than I’ve already said - in fact less. Well I’m happy to know that I’m better than “chat gbt” but it really doesn’t speak well of your understanding. The more so since your claims of education and research imply that you should have known it already.
None of this is empirical. You are counting stuff with zero consideration for how well or how poorly that represents experimental data.
No, they will not. Firstly, it’s a nonsensical question to ask whether or not some quantity is real. Is the number of stars in the sky real? What does that even mean? What would it mean for it not to be? Identifying some observable and giving it a name does not solve any problem anybody actually had, nor does it on its own generate predictive models. What you mean is whether or not there is some behaviour this “genetic entropy” really, measurably, undergoes. Whether it, say, only ever increases with time given some reasonable boundary conditions, as its thermodynamic counterpart does. That is to say, whether there is some measurable quantity one may call “genetic entropy” that behaves in some systematic way one can utilize to render testable predictions. But no amount of estimating favoured or total possibilities of some mathematical abstraction will answer that question. Measurement will.
I would love to see that logical argument actually formulated. Would you, kindly?
Please, if you are going to talk maths, be precise. “… multiplying each position given the independence of the events” is not a theorem. There is no such thing as a “fundamental proof of” that. What you mean is the “proof that the probability of independent events coinciding is the product of the probabilities of each individual event”. And that is a consequence of the definition of statistical independence. It has nothing to do with the counting measure or combinatorics more broadly. You either lied when you said you studied this, and genuinely do not understand it, or you did actually study it, know better, and lie about it anyway. Neither situation will alter how it is, though.
A genetic fallacy is saying that some statement is incorrect because of its source (rather than some commonly better respected metric for a statement’s accuracy, such as the state of the relevant evidence base). Pointing out an inaccurate description/classification of a source is not a genetic fallacy.
I realized what he was when I read his literalist polemic on the patriarch lifespan decay curve, where he claimed that no ancient Hebrew could have made up the ages if GE were not authentic, as they lacked the mathematical skills. Aside from the cherry picked data points, of course any reasonably ordered set can be regressed.
Add to that his misleading influenza paper.
By his own writings, the man is entirely motivated, whether for reasons heartfelt or ulterior, by creationist and specifically YEC zealotry. Do you really expect that Sanford would code a program and report back that “after running this totally objective and fair program, I am convinced that evolution is true and all my work on GE is bunk”? Of course he is working backwards from the desired result. More like Capone’s Accountant than Mendel’s Accountant.
Yes for Sanford is much more concerned by his moral integrity than his scientific reputation. But instead of accusing him of dishonesty, you would better explain what it is precisely that makes MA a non objective and unfair program. As a first step, you may want to review the paper I refer to above and spot where it is dishonest.
Even if it is sophistry, at least it’s true. Unlike your claim that figures not generated by Mendel’s accountant can show how it allows fitness to increase - a claim that you haven’t retracted.
Fig 3a shows no mutations with a positive fitness effect, and the accompanying text[1] is about high impact deleterious mutations.
Meanwhile, Fig 1 contradicts the text that mentions Fig 3, because it shows you can set a maximum fitness effect for beneficial mutations that is the same as the minimum fitness effect for major deleterious mutations.
Do you actually read the papers you cite?
“Response of the fitness-effect distribution function to changes in the fraction of “high impact” mutations (0.0001 to 0.1)”↩︎
It will give us a rough estimate based on current empirical data that continues to improve.
Yes selection is part of the variation reduction filter and can add variations to current alleles to a population.
Chat Gbt is better than we are in the way it can learn and is impartial. Unlike humans it cannot think outside the available ideas unless you help it. The relationship between probability and combinatorics appears to be deeper than chat gbt’s original response and your original assertions. Discovering this will take more thought.
If we think about a single position of a sequence as either the numerator or denominator of a probability ratio since it represents a population then it makes sense that you can multiply these ratios given the rules of probability of independent events.
This is gibberish. Noone here, including yourself, has any clue what “a single position of a sequence” means. Noone here, including yourself, understands how to think of “a single position of a sequence” as either a “numerator or denominator of a probability ratio”, and “probability ratio” might as well be “flux capacitor”, for all the technical meaning it bears. Noone here, including yourself, could articulate exactly in what way “a single position of a sequence… represents a population” either.
What “ratios”? You said that if one were to think of a single position of a sequence as either a numerator or a denominator of a ratio. Oh, sure, I guess you can pull random numbers out of a hat, and arrange them into ratios like that, and sure enough one can multiply them all one wants. However, what sense any of it makes is completely lost on everybody here, including yourself.
Well, until someone can demonstrate that we are talking of independent events, I say none of the results so (i.e. by making it right up, since for the life of me I have no clue how else to interpret this incoherent babble) obtained necessarily mean anything in practice.
And? It certainly won’t be anywhere near adequate to us-port your claims about searches.
And yet you did not mention it - only DNA repair and drift.
Not really since it can’t actually learn and it doesn’t distinguish between truth and falsehood - which is a sort of “impartiality” that is less than helpful. And I really think that your “thought” will just result in you spouting nonsense again.
