Thanks for the comments, @Rope. Just so things are clear in this discussion here:
- Axe’s estimate for the fraction of functional sequences is wrong. Period.
- Axe is wrong. But even if not, his work does nothing to change the vast range of estimates that existed circa 2007 (or 2004). This range goes from “pretty improbable” to “almost certain”. Basically, probabilities from 0 to 1, inclusive.
- Lots has been done since then. We had a very insightful discussion here on PS that further eroded what little confidence might be had in Axe’s work. Especially relevant was the realization that one of Denton’s proposals - that proteins are in some way “platonic forms” (sort of a molecular version of “laws of form”) - is wrong. (I say realization as if it only occurred to discussants on PS at that time. This is really not true - this has been known for decades before PS ever came along.)