Hunt's 2007 Critique of Axe

What is the design principle behind a rube-golberg machine?

1 Like

There are also a nearly infinite number of combinations of protein sequences and protein functions that can produce a network of interdependent proteins and small molecules. Testing for just one of those functions ignores all of the other possible functions.


In what way is it helpful?


If you have some functional proteins in a cell there is not unlimited possibilities with new emerging proteins. They must bind with the existing proteins. In Axe’s case if the bacterial protein did not bind penicillin it would die. The “any protein will do” is not relevant to living organisms.

The sequential nature of DNA and proteins is a very large problem for evolution to explain.

For example I was doing research for trying to find the cause of why low levels of vitamin D were causing cancer initiation. It turned out that vitamin D was down regulating cell division by up regulating the destruction mechanism of the ubiquitin system for a specific transcriptional molecule.

When I was asked to research why high blood levels vitamin d were reducing risk of metastasis I used design as a working hypothesis. I used a common mechanism as a working assumption and this turned out to be a tremendous short cut.

1 Like

21 posts were split to a new topic: Mechanistic Reasoning is Not Design-Based Reasoning

Did Axe determine if the proteins he was studying bound to any other proteins in those organisms?

I never said that “any protein will do”, so I am curious why you keep saying that. What I am saying is that mutations can produce many different functions. The problem with Axe’s study is that he just looks at one function, and from that one study proclaims that function of any kind is improbable. Do you see the problem?

1 Like

Yes the protein sequence that he mutated was half of total enzyme used. So post translational protein to protein binding occurred.

I don’t think this is Axe’s claim and Hunt’s main critique was not of Axe’s work but with the ID claim you mentioned. What Hunt’s paper showed is a range of probabilistic resources to find specific functions. Evolution does have to find specific binding targets so this study is problematic.

The sequence is a big problem for any blind and unguided claim.

I don’t remember Axe assaying for protein-protein interactions for the different mutations, but I should probably go back and check.

First, I don’t see why evolution has to find specific binding targets since there are tons of bacterial strains out there without a beta-lactamase gene. Humans don’t have a beta-lactamase gene and we do fine without it. Therefore, beta-lactamase doesn’t have to evolve.

How so?

1 Like

If it is of any length greater the 20 amino acids the number of ways to arrange it is very large. There is no logical way to arrange it functionally other then conscious intelligence.

I have yet to see how this logical argument works.

1 Like

Create 50 functional telephone numbers other then having conscious knowledge of the sequences or numbers.

How is that a logical argument for protein function?

As I understand things, it is not unusual for a protein to have some function. It doesn’t need to be the exact function for which we observe in its current form, it only have to be enough function to get the job done. That seems to be highly relevant, because we expect the interdependence will evolve over time.


Protein function like telephone function lives in a sequence. It is in an amino acid sequence versus a number sequence. What do we know that can create a functional sequence?

On what basis do you expect this? The assumption that protein evolution is true.

To get a cell to divide reliably you need optimized sequences. This is a minimum criteria for a living organism.

That seems transparently false.

1 Like


I just posted something on this. I will try to find it but definitely said you can skip steps and get steps out of sequence.

Actually, in this case the protein-coding gene most definitely arose from scratch. Via well-understood processes. And starting with sequences known for a fact to not be protein-coding.