The Jukes-Cantor and Kimura two-parameter models differ in that The parameters of the Jukes-Cantor model represent mutation rates, whereas those of the Kimura two-parameter model represent substitution rates. The Jukes-Cantor model assumes that all base substitutions occur at the same rate, whereas the Kimura two-parameter model assumes that substitutions between purines and pyrimidines occur with a lower probability than substitutions of one purine for another or one pyrimidine for another. The Jukes-Cantor model applies a parsimony criterion to determine the tree that best fits a set of character data, whereas the Kimura two-parameter model evaluates the posterior probability distribution to choose the best tree