PAM250 extrapolations

Note: these PAM250 extrapolations have not been rounded to the nearest integer. They have been done with Failed to parse (Cannot write to or create math temp directory): 10*log_{10}

instead of Failed to parse (Cannot write to or create math temp directory): 2*log_2
like the other matrices, because Benner uses that C.

[edit] 1 PAM250 extrapolation sums

Sums are of the upper triangular extrapolated PAM250 log odds matrix. Includes diagonals.

[edit] 2 PAM250 class differences

[edit] 3 PAM250 expected values

Had to check this--expected values should all be negative, and the positive sum for D40 was worrying me.

[edit] 4 Comparison with Genetic Code matrix

Disordered gets much closer to the genetic code matrix over time...but they are all very very different from the genetic code matrix still.

For comparison of how different that is, the differences between disorder and order PAM250s are less, that is, they are more similar to each other than they are to the genetic code matrix:

Despite *not* having any steady increase/decrease pattern from 85% to 40% in the difference between order and disorder, disorder compared to the genetic code matrix steadily decreases as divergence increases.

This might be because of high diagonal values. Table of nondiagonal differences between PAM250s and GCM:

[edit] 4.1 log2 Test

All the GC comparisons had to be done with Failed to parse (Cannot write to or create math temp directory): 10log_{10}

values, but what about when we use the Failed to parse (Cannot write to or create math temp directory): 2log_{2}
substitution matrices?