Keep in mind I am not saying anything about the veracity of his paper except it was published, is in the news, and the author considers it ID research and publication.
I think this is the problematic point. Does the fact that the author considers this ID research actually make it in any way supportive of ID. The Behe and Snoke paper has already been mentioned and despite having been pushed as supportive of ID it still showed that irreducibly complex features could arise, and that was with some highly restrictive assumptions in the modeling.
Axe doesn't actually seem to be estimating the likelihood of anything by evolution,but rather that of any randomly generated protein domain having a specific function.
The actual linkage of this to problems with evolution or support with ID is tenuous. Axe writes ...
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.
This argument seems to be that since we can't necessarily recreate a functional ancestral protein just by cutting and pasting elements of homologous proteins together then domain shuffling and other common genetic mechanisms may not be sufficient to generate the diversity we see in protein structure. It has very little to do with the figure Axe estimates, since that figure was for the proportion of all possible sequences of a particular length which could adopt very specific stable conformation and confer function.
Axe himself writes ...
At first glance, it seems implausible that natural sequences could diverge through a space where function is represented so sparsely. How, for example, can we account for the substantial diversity among the large-domain homologues (Figure 5) if randomly altered sequences have such slim prospects of retaining function? The answer follows from the fact that functional sequences are not distributed uniformly through sequence space. A random change to a functional sequence actually has a good chance of leaving function undisturbed if very few positions are affected. As estimated above, the likelihood of a signature-compliant substitution in the large-domain reference sequence producing a comparably functional variant is about 38%. Since 70% of the ~1000 possible non-synonymous base changes to the reference coding sequence produce signature-compliant substitutions, about one in four random single-residue changes are functionally neutral. The proportion would be somewhat lower under conditions requiring a higher level of function (such as those under which neutral drift normally occurs) but not so low as to preclude progressive sequence divergence by gradual accumulation of point mutations.
So where is the support for ID? Axe generates a
really big number but doesn't tie it in to evolution. His argument about domain shuffling seems almost entirely divorced from his calculations of the proportion of stable functional large domains in the whole set of possible sequences of large domain length.
I'm also not sure of his phrasing as being concerned with 'the rarity of sequences expected to form working enzymes' since the paper discusses 'the proportion of all sequences of large-domain length that perform the specified function by means of any tertiary fold'. There is a world of difference between forming a specific functional fold and forming any functional fold. The only reasoning I can see for this is that he may be shaving a few orders of magnitude off as a way of encompassing all 'working enzymes' and basing this reduction on the estimates of how many extant fold types are needed to cover all biological functions. This fails to take into account the fact that this does not represent all possible functional folds making the estimate omly about the probability of randomly coming up with a long domain sequence forming a stable fold which confers function
and corresponds to one already extant.
Axe also gives only the highest range of the reported figure (one in 10
77), using his lower estimate (one in 10
53) you could knock a couple of trillion off it.
Doug Axe's papers always strike me as being pretty good science,although I'm not sure how useful Stylus will turn out to be it is certainly interesting, but the support for ID is only there if you squint and want really hard to see it.
TTFN,
WK
Edited by Wounded King, : No reason given.