The term “functional coherence” is used in this paper.
Wow Bill you found the term used in some obscure and irrelevant scientific paper? What a remarkable new development in this deeply scientific drive-by of yours. 
2 Likes
First “dramatic nesting”, now “functional coherence”. How long can this list get?
3 Likes
It is a fathomless abyss.
1 Like
But it’s not mentioned in any of the papers discussing FI, so it has nothing to do with FI.
What we have is an ID method that says randomly changing a sequence will add information to it. Why they think this can’t happen with known and observed natural processes is beyond me. Can you explain it?
3 Likes
Apologies. I was adopting the rôle of ingénu. Of course I’m aware that nobody has yet calculated the functional information of anything. Carry on.
1 Like
Wrong. The challenge doesn’t assume all 500 bits are randomly assigned in a single trial, it allows multiple trials, selection and large populations !
Your scenario here doesn’t model the emergence of a complex functional sequence from an unrelated state by blind natural processes, not at all.
If we use Gpuccio’s method that scenario does produce the emergence of a complex functional sequence as measured by the increase in FI.
Or . . .
The method used for measuring FI does not do what you claim it does. If the method for measuring FI does reliably measure the emergence of complex functional sequence then it shouldn’t show an increase in FI for randomly changing random sequence. The problem for you and Gpuccio is that it does show an increase in FI when you randomly change a random starting sequence.
1 Like
I think they did in Szostak and Hazen (et al.)'s paper in 2007. It is meaningful, just not a measure of how unevolvable anything is.
1 Like
Why would we want to model evolution in a way that we already know it rarely uses?
The use of a threshold of functional information as a supposed impossibility is based on the amount of the FI. Not anything about “complex”.
1 Like
I mean that nobody has calculated the functional information in any DNA or protein sequence. If I recall, they used a much simpler and artificial sort of system.
I disagree. Complexity has something to do with high FI. A bio sequence with high FI is necessarily complex (in the sense of improbable). But of course, complexity is not sufficient for a bio sequence to exhibit high FI, for it must also be functional.
That’s not what “complex” means. Not to anybody. Szostak’s definition says nothing about complexity, though it does refer, sort of, to improbability, i.e. as being far out on the right tail of a distribution. And that distribution is in fact one of functionality. But I wouldn’t say that there’s any reason to expect that the most complex sequences in that distribution (by whatever definition) are also the most functional.
3 Likes
Agree.
Only if you have landed on an island of function. But what if islands of function for complex systems such as the ATP synthase are rare in sequence space? How do you land on such islands from an unrelated state?
Agreed.
This is a very complex contraption:
… but I don’t think many would consider it to be the most functional means of achieving its purpose.
2 Likes
In any case the challenge has been met. A genome sequence is a functional sequence, and depending on the ratio of deleterious to beneficial mutations, 500 bits of FI has already evolved in the LTEE (if the ratio is 1:106). Even if one disagrees with that ratio and thinks it’s higher than one in a million, it’s obvious natural selection can continue to just increase FI basically indefinitely.
By mutating incrementally under selection from the unrelated state?
1 Like
Your math required 26 beneficial mutations in a row (assuming 1 in 1 million frequency of beneficial mutations). When did this occur?
This also assumes that function=fitness which is a suspect definition of function. Where FI is working in the cell function is generating ATP or allowing the cell to divide and 500 bits of FI is a small protein which works with other proteins.
During the course of the LTEE.
The ability to survive and reproduce (cell division is of course part of reproduction) is a biological function (some would say it is the function of life, or at least one of the most important functions of life.) Fitness is a way to measure that function.
FI can in principle be applied to any function you can imagine. There is no requirement it has to be specifically applied to the function of an individual gene or protein. In Hazen & Szostak 2007 they apply it to Avida populations ability to perform certain logical operations too.
You define the ability of a cell to survive and reproduce as the function. Ok, why not. But you realize that in your exemple of the LTEE experiment, the function is already there, don’t you? So we are not in the case where a new complex function emerges from an unrelated state, are we?