Sauer's laboratory, in order to answer questions about protein structure that interested them, took the genes for several viral proteins, systematically took out small pieces of them (corresponding to instructions for three amino acids at a time) and inserted altered pieces back in the genes. They did this, three amino acids 'codons' at a time, for the whole length of the gene. By clever manipulation of the altered pieces they were able to screen codons for all twenty amino acids at each position of the protein. This is like trying all 26 letters of the alphabet in turn at each position of a word. The altered genes were then placed in bacteria, which read the DNA code and produced chains of amino acids from them. It turns out that bacteria quickly destroy proteins that are not folded, so Sauer's group looked for the altered proteins that were not destroyed. By determining their sequences they could tell which amino acids in a given position were compatible with producing a folded, functional protein. And what did they see?
.
.
.
In some positions of the protein Sauer's group saw that a great deal of amino acid diversity could be tolerated. Up to 15 of the twenty amino acids could occur at some positions and still yield a functional, folded protein. However, at other positions in the amino acid sequence very little diversity could be tolerated. Many positions could accomodate only 3 or 4 different amino acids. Other positions had an absolute requirement for a particular amino acid; this means that if, say, a P does not appear at position 78 of a given protein the protein will not fold regardless of the proximity of the rest of the sequence to the natural protein. In terms of our sentence analogy, this is like saying that, yes, all vowels are interchangeable, but that if the last `r' is changed to any other letter, such as 's' ("Drop the anchor in one hous"), the protein sentence is no longer understandable.

.
.
.

From the actual experimental results of Sauer's group it can easily be calculated that the odds of finding a folded protein are about 1 in 10 to the 65 power (6).
.
.
.
It is important to realize that Sauer's and Yockey's results hold whether or not the system can replicate and is subject to Darwinian selection. The odds against finding a new functional protein structure remain astronomical in either case. This is because Darwinian selection can only discriminate based on function and, with the exception of those found in living organisms, virtually all protein sequences are functionless. An amino acid sequence can be replicated and mutated in living organisms till the cows come home and the odds are still 1 in 10&% that a new functional protein class will be produced.