2. Comment #448567 by healthphysicist
The problems with Axe's work have been discussed here:

http://pandasthumb.org/archives/2007/01/92-second-st-fa.html

Thanks for that link.

Unfortunately people here need to read and understand it. I say "Unfortunately" because it was a complex biochemical experiment. I say "need" because this was carried out by a PhD level researcher funded by the Discovery Institute. He did do his 8 years of school. You can't dismiss this with handwaving, you need to understand the experiment and its flaws.

I'm not a biochemist so I am probably talking bollocks, but ...

The 4 regions of 10 amino acid sequences that were mutated were separated in the 3d structure. Could "complimentary" mutations in adjacent regions increase the number of functional sequences?

For example, suppose a "protein" is folded:-

x-
-x
x-
-x

where a "x" must have a "-" on the row above and below, and visa versa.

Suppose we mutate rows 2 and 4. Of the 16 possible mutations only 1 is "functional". This gives an overall estimate of 1 in 256 sequences is functional.

However, if everything can mutate, then

-x
x-
-x
x-

is also valid, which doubles the functional sequences to 2 in 256.

Maybe this goes some way to explain the gap between 10^-10 and 10^-70 in "forward" and "reverse" experiments.