Context-dependent substitution matrices for coding regions of angiosperm cpDNA

The dataset contains 4x4 substitution matrices generated from fourfold
degenerate sites within coding regions of flowering plant chloroplast DNA.
Substitution data was generated from multiple sets of 3 taxa sequence
comparisons, one taxon serving as the outgroup to root the substitution.
There is one matrix for each of the possible 192 tetranucleotide contexts
consisting of the two bases immediately flanking the substitution on the
5' side and the two bases immediately flanking the substitution on
the 3' side. These are count matrices but can be converted to Markov
transition matrices. The analysis presented in the paper examines
variation in substitution dynamics as a function of context.