A strategy for sequence phylogeny research.

Minimal mutation trees, and almost minimal trees, are constructed from two data sets, one of phenylalanine tRNA sequences, and the other of 5S RNA sequences, from a diverse range of organisms. The two sets of results are mutually consistent. Trees representing previous evolutionary hypotheses are compared using a total weighted mutational distance criterion. The importance of sequence data from relatively little-studed phylogenetic lines is stressed. A procedure is illustrated which circumvents the computational difficulty of evaluating the astronomically large number of possible trees, without resorting to suboptimal methods.