Molecular data offer the possibility of helping to resolve relationships among distantly related organism. Nucleotide sequence data have the potential to be especially informative. But one difficulty is that there are only four character states (representing the four different nucleotides) at any one position in the sequence. If the molecule is changing rapidly or the time since divergence is long, multiple substitutions at any particular position may obscure information about relationships of the organisms. If we knew which positions were most likely to have sustained multiple substitutions we might give that information less weight in an analysis of relationships. This project will test one method of determining which positions are more likely to have sustained multiple substitutions. We will obtain sequence data of a particular chloroplast gene for genera within one family of flowering plants and determine which nucleotide positions vary within that family. We will then obtain sequence data of that same gene from a number of divergent angiosperm families for which we have a reasonable idea of the true relationships. We will use the information on variation in nucleotide sequence within one family to help decide which nucleotide positions provide accurate information on the relationships of the divergent angiosperms. The similarity of results of analyses of relationships using information on the likelihood of multiple substitution with the true relationships will be used to evaluate the efficacy of the method.
