Homoplasy estimators with a comparison across data types

Homoplasy - similarity of structures for reasons other than shared common ancestry. Understanding homoplasy is essential to understanding evolutionary processes. Yet rigorous investigations of the topic are rare in the literature. Perhaps this is due to the difficulty in defining homoplasy. It is impossible to fully separate the notion of homoplasy from its sister homology. What is meant by homoplasy is dependent upon the definition adopted of homology. With the rise of phylogenetic reconstructions, the study of homoplasy has shifted from a biological focus to a phylogenetic focus. Its utility being the fitness criterion for most phylogenetic studies assuming parsimony. Thus if separate data sets are utilized in phylogenetic studies each is subject to their own caveats affecting the level of estimated homoplasy. In this work, I seek to explore the interface of homoplasy and mixed data sets. Specifically, this work does four things. First, reviews current estimators of homoplasy. Second, discusses mixed different data types, combining data types and utility of homoplaseous characters. Third, asks if there are consistent differences in estimated homoplasy between morphological and molecular data types. Fourth, examines sequence data sets as a chimera of component data sets and asks if there are significant differences in estimated homoplasy between these component data sets. Some of these issues have been addressed in the literature. However, each section provides a positive contribution to the field. Occasionally, results are opposite those in the literature. Specifically, I have shown (a) that there are significant differences in estimated homoplasy between morphological and molecular data, and (b) that there is no significant difference in estimated homoplasy between the third codon position and the combination of the first and second positions. Other results are equally important though possibly not as unexpected. Finally, the results contained herein are relevant to a number of topics in evolutionary studies. Including the relationship between different data types, inherent levels of homoplasy and whether or not to combine different data types among others