I did no such thing and you know it.
That’s objectively false, Bill.
This is called a quote mine. Why do you feel the need to do this?
And there we have it, you think you’re right by default and I have to prove you wrong. No, you have to demonstrate that there are no other sequences that meet the minimum threshold for function. Demonstrate, with evidence and experiment. Not insist, offer opinions, or merely rationalize. DEMONSTRATE.
2 Likes
What context would change the meaning?
Yes we do know this Rum. You can calculate this by the size of the sequence and the possible substitutions that can get you say 150bits or more. The restriction is less than all AA’s but 2 having substitutability.
Not a chance of that. It would require work and acceptance of the fact that digging into the evidence (such as the >5000 papers on catalytic antibodies) would show that functional sequence space is very large. Bill’s afraid to measure the numerator of the ratio he claims is so small.
How does that tell you there are no other sequences in that space that can perform the function?
No, @gpuccio’s calculation of FI requires you to know the size of the target space. His estimate is laughable, but at least he’s trying. You don’t have a clue and you are afraid to look.
3 Likes
It’s more that he deeply wants to believe what Gpuccio is doing is valid, so he’s highly motivated to offer up any rationalization and excuse he can think of to defend it even when they don’t logically work.
1 Like
It doesn’t. It tells you the window of evolvability is small. Add 400 million years of 90% preservation and the evidence is compelling against the evolution of this protein. If there was 50% substitutability would you expect to get this level of preservation? We see saturation in the DNA changes but strong restriction in the AA changes.
Dr. Mantha, first as a total aside, I’m interested in learning some Physical Chemistry. I was told since I have some physics background (like Quantum Mechanics, Thermodyanmics and Statistical Mechanics), I should enjoy P-Chem. Are there books you could suggest. I figured you’d be the guy to ask.
What function means is in they eye of the beholder. It’s relatively easy for example for almost any human to make a paper weight or some sort of primitive hammer, it doesn’t me they can make a computer chip. The issue is the difficulty of first conceiving a design and fabricating it.
Some of the examples that physical chemist and cellular biologist Change Tan pointed out for origin of life are the necessary, but not sufficient, chemical requirements to make a functional DNA/protein system. That is to say, one could conceive of the necessary requirements rather easily, but from a chemistry perspective in a pre-biotic environment, it is astronomically improbable.
The problem she highlighted is the homochirality in nucleotides and amino acids and the “homolinkages” happening in solvent. These aren’t even necessarily functional sequences, but just the general policy that is necessary but not sufficient to create a functional sequence.
By “functional” I mean something that can be integrated. For example one could argue my D-amino acid protein isn’t functional because there is no cell that exists that could use it, but in principle there could be a cell that could be constructed that could use it.
But what Dr. Tan was pointing out was the simple-to-understand design of homolinkage and homochirality. The problem of homolinkage is especially bad for nucleotides being assembled into a DNA genome from a prebiotic environment. She points out the many hoops that biotech companies like Blue Herron must go through to make linkages.
The point is, much like 500 fair coins being all heads, it’s easy to conceive of and implement. But in a pre-biotic environment where the “coins” are the linkages and chiral parts, there is no one around to do the flipping to heads, so to speak. Therefore, Dr. Tan, who was an atheist, eventually became a Christian creationist.
The design of a protein of modest complexity is another challenge, much like making a lock that will fit around a given key. There are a lot of mechanical connections that need to take place, and like building parts that connect right, that’s the problem with building proteins complexes like this:
You’ll see the KRAB KZNF protein in the diagaram above. I often show this example of such a KRAB ZNF protein sequence here:
We’re trying to build AI that can properly modify the rows in the above protein with the colored columns, but it doesn’t work so well even though the fold of each zinc finger is trival.
That’s a good way of putting it.
If he believed, he’d be eager to dig into any evidence he could find. Instead, he projects his complete lack of faith onto those of us who literally produce the evidence he is afraid to examine.
1 Like
Then how can you claim to know how many sequences meet the minimal threshold for function? You’d need to know that to calculate FI.
If, as you just admitted, you don’t know how many sequences meet the minimum threshold for function, then you can’t calculate FI.
That’s it, we’re done. Game over.
It tells you the window of evolvability is small.
How does it do that? You have no idea how many such windows there are by your own admittance. Even if each window is small, you’d need to know how many there are and how far apart they are. Do they overlap windows to other functions that they could evolve from, and how do you know? Where does the “search” start and how do you know?
Add 400 million years of 90% preservation and the evidence is compelling against the evolution of this protein.
All that shows is further change is deleterious, it says nothing about how the present state came to be. That’s like saying the foundation for a house could not have been poured because there’s a house in the way. Uhm, the house came later.
All your excuses and rationalizations are bad, they don’t work. They fall apart INSTANTLY on even cursory analysis.
2 Likes
It’s because the arguments against him are not strong. You are resorting to “how do you know” arguments which is nothing more than a burden shift.
We can estimate it Rum. The “how do you know” is a burden shift. Science is not about absolute proof.
But you HAVE the burden by default when you’re claiming to have calculated the FI for a particular protein. You TAKE ON that burden in doing that calculation. You now have to justify the values you plug in to the calculation Bill, and you just admitted you can’t do that.
My work here is done.
1 Like
Doing the calculation requires numbers to plug in to the equation. You have to get those numbers from somewhere. But for the calculation to work, you need to know what the numbers actually are. That’s just how equations work. And you admit that you don’t know what the numbers actually are.
Then you can’t do the calculation, and you can’t claim to know the FI of a known protein. Thus, you can’t claim to know it contains 500 bits. Thus you can’t claim to know it couldn’t have evolved.
1 Like
Why don’t you explain what you mean by “how do you know”.? There is a burden to supply and calculation that estimates FI and that has been done. You have no real criticism other than to invoke the unknown.
In doing the calculation, you are claiming to know the unknown. When you plug in a number for the minimum threshold for function, you are claiming to know there are no other sequences than the number you plug in. This claim is yours, when you make it you have a burden to show that the number is correct, to show how you have determined that there are no other sequences than those you have included in your calculation.
Obviously you have to justify that knowledge claim. How do you know that? You could say “I have done experiments, I have tested all possible sequences” for example. Naturally for a large protein that claim would be extremely implausible, so you haven’t done that. But then… what work have you done that tells you there are no other sequences than the ones you included in your calculation? None, you’ve done none at all.
You have offered some arguments to think that the number of sequences is restricted. I don’t have any obections to that in the broad sense that, yes, the more different specific interactions a protein has to do to work, the more restricted the number of sequences required to perform the function is going to be. But, by how much is it restricted? What is the magnitude of the restriction? You have to tell me actual numbers here and how you got them. Like your math teacher would say, show your work.
1 Like
Let’s see if this helps you to understand the problem you face.
Suppose I have before me a bag of marbles, each of which is either black or white.
The bag contains exactly 5 black marbles.
What percentage of the marbles in the bag are black?