BLOSUM62 matrix: M to L has higher probability score than M to M

[ywlim]

I noticed in BLOSUM62_PROB.Rdata, M to L substitution has higher score than M to M. This is also true in the BLOSUM62 matrix from http://www.cs.cornell.edu/courses/cs626/2003sp/blosum62qij.txt.

For example, I was not able to generate the same image shown in the tutorial using the same peptide sequence.

"FIYRHMFCV" %>% pep_plot_images

Below is the generated image. Notice the score for substituting M_6 with L is the highest.