Phylogeny determination using dynamically weighted parsimony method.

Publisher Summary This chapter presents various methods for assigning two kinds of weights plus a method for evaluating trees given the weights. A character is a nucleotide position, and a character state is the particular nucleotide in that position. There are two kinds of weighting. The first is character weighting in which the weight or importance to be given to any position may differ from that given to other positions. These weights are stored in a character weight vector, sometimes referred to hereafter simply as the C vector. Dynamic weighting means altering the values of the C vector and/or the T matrix as a consequence of the results of the previous pass. With these new sets of weights, researchers make another pass repeating the search for the best tree. The user has the option of starting subsequent passes using either the seed tree or the best tree of the previous pass. Whether the result is more likely correct than when unweighted parsimony is used has not been proved. It ought to be true in principle, but the improvement might be marginal.