An acquaintance provided me with this article1. I can't understand for sure what it is about.

My acquaintance said that it proves that time for generation of even the simplest proteins is on a larger timescale than evolution history (billions of years).

I'm asking you, is that a valid conclusion to be made from this paper? What is in this article that might drive someone to that conclusion?

  1. Axe,D.D. (2004) Estimating the prevalence of protein sequences adopting functional enzyme folds. J. Mol. Biol., 341, 1295–315. http://www.ncbi.nlm.nih.gov/pubmed/15321723
  • $\begingroup$ In addition. As i understand, this man considered linear exhaustive search of nucleotides. As i understand it is wrong, because genetics and evolution algorithms are optimized exhaustive search. $\endgroup$
    – dshulgin
    Sep 18, 2016 at 19:11
  • $\begingroup$ I appreciate the way you phrased your question in terms of appropriateness, and the good will of the person that gave you an answer, but this is really not an appropriate question. The main reason is that it is not comprehensible as it stands. First one is required to go elsewhere to read a paper (which happens to be behind a paywall), and second the paper (at least the summary) makes no reference to evolutionary time. Thus, we do not know what argument your acquaintance is making. So, with reluctance, I am voting to close the question unless you can edit it to make it fully self-contained. $\endgroup$
    – David
    Sep 19, 2016 at 21:33
  • 1
    $\begingroup$ @James — I have no objection to Elsevier charging for what is there's. Your language is another matter. Even if the article were not behind a paywall the question is unsuitable for SE Biology as it requires one to read a paper and then imagine a third party's attitude to it. If you ask similar types of questions they too should be closed. The guidelines on writing questions are quite clear. $\endgroup$
    – David
    Sep 20, 2016 at 6:32
  • $\begingroup$ @David I've asked similar questions about skeptically analyzing conclusions drawn from articles. It certainly shouldn't be closed because of Elseviers shameless paywalls. Often abstracts don't include some observations, particularly long technical papers with advanced and nuanced methods like this. +1 good question. $\endgroup$
    – James
    Sep 20, 2016 at 7:41
  • 1
    $\begingroup$ You might find some useful analysis of the paper here: pandasthumb.org/archives/2007/01/92-second-st-fa.html $\endgroup$
    – swbarnes2
    Sep 23, 2016 at 20:23

3 Answers 3


I think your acquaintance is trying to fit real science to some of his personal beliefs (that are obviously wrong).

If you read the article you'll see that it's not about evolution at all, but about protein folding and what proportion of possible sequences gives a working protein. It turns out random sequences are not that likely to fold, which leads to some conclusions about the probable mutational pathways to folded proteins.

The argument of your friend is sort of analogous to the infinite random monkey theorem. If you put a monkey behind a typewriter and wait long enough, eventually he'll write the complete works of Shakespeare. This doesn't mean we waited billions of years for this monkey to type it up, because the world doesn't work like that. There was just a guy with a brain.

In case of the proteins, evolution doesn't work by generating newly (random) 300 residue proteins that work 1 time in a billion. Probably in one generation a copy of the protein is made with the same function, and during many generations the function of this protein drifts towards a new function, while all the intermediate proteins stay folded because they're not that different.

  • $\begingroup$ For some background on the author of the paper (Douglas Axe) see for example rationalwiki.org/wiki/Douglas_Axe. My personal opinion is that the paper looks interesting from a quick scan, but there may be some agenda behind it. $\endgroup$
    – gilleain
    Sep 19, 2016 at 10:57
  • 1
    $\begingroup$ Also from the wiki: Axe himself says his work does not support intelligent design. So there might be a conflict of interest, but the data in the paper doesn't prove anything regarding evolution anyway. $\endgroup$
    – VonBeche
    Sep 19, 2016 at 11:03

Conclusions from the paper.

The finally line in the article should actually be read first as it clearly spells out where the paper fits in the scientific literature.

So, if re-creating a fold by ordered assembly of sections of sequences that already adopt that fold is not a simple matter, generating new folds from parts of old ones may be much less feasible than has been supposed.

A popular hypothesis at the time was that some folds emerged at some point or another from substantially different folds.

This paper used advanced statistical analysis (well, they're convoluted, not particularly advanced...) using a very simplified and uncertain model to show that the chimera theory of fold genesis is unlikely. It is a critique of that model at the time, and does not seek to show how long it took for evolution to come up with the fold.

This is an incredibly important question to answer. This paper was probably accepted because it shows a genuine attempt to quantify the probability of a general hypothesis. However, almost every step of their method introduces over-simplification, arbitrary thresholds, and convolution of the biology. As I see it, the only thing the paper can say for certain is "that in a small subset of simple folds, with a tonne of caveats and conditions, it might be less likely than other people think that the fold formed that way".

What your friend is drawing from this.

It would be incredibly difficult, many philosophers would argue impossible, to prove how folds were formed. But to build up lots of different scientific hypothesis and disprove them as much as possible is more achievable. That, broadly speaking, is the scientific method and this paper generally adheres to that line of thought.

This paper is saying "The current hypothesis surely can't be the right method! Look how improbable it is." and your friend has drawn from it "A scientist said more time would be needed, which proves evolution needed longer than it had". A simple misunderstanding; a scientist said a model had problems, a layman thought this meant the problems were the truth. The paper is not aiming to prove anything (The author may be especially careful about doing this given his close ties to intelligent design movements), but I think your friend might be incorrectly using the paper to prove something!


The study by Axe (a creationist) is fatally flawed. See these comments on it by Arthur Kent, a biologist at the University of Kentucky. To summarize: Axe did not study the native form of the protein he was looking at (TEM-1 penicillinase), but specifically, a variant that both had lower activity and was especially temperature-sensitive. His results, then, can't be generalized to even the native form of the enzyme under investigation, let alone any other enzyme. As Kent points out, more recent techniques suggest a substantially larger proportion of functional variants than Axe's method suggested. In any case, experimental and observational evolution has settled this debate: we see proteins with novel functions constantly evolving from pre-existing proteins all around us (how do you think new bacteria are constantly emerging which can digest all sort of novel artificial synthetic plastics never before seen in nature? how do you think viruses switch hosts?), and so it is 'out of question' as to whether there is sufficient time for new proteins to emerge. We can see it happening ... all the time!


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .