A New Measure Of Information Discrepancy And Its Application To Biologic Evolution

Since THE ORIGIN OF SPECIES of Darwin was published in 1859, the theory of biologic evolution has been one of the most contributions to the development of natural science and natural philosophy. Evolutionism is the most important theory in life science. With the development of molecular biology, the research of evolution has been pushed to the molecular standard. So the research of phylogenetics on the standard of molecule is the important part of bioinformatics. In this paper, the main work is about FDOD and its application to biologic evolution, it is summarized as follows:1. Multiple sequence comparison is a basic problem for molecular biology. In this paper, a new information theory method, which is noted by FDOD, is used to infer the phylogeny of SARS and three known coronaviral groups based on four different protein sequences. The phylogenetic tree agrees with the existing biological analysis. FDOD only depends on the original information about sequences, no other subjective factors are involved. Furthermore, the process of computation is simple and quick, the result shows that the FDOD approach is a useful tool in studies of molecular evolution.2. With the accomplishment of more and more microbial genome sequencing, people have begun to study phylogenetics on the standard of whole genome, In this paper, FDOD method is used to infer the phylogeny of whole proteome microbial organisms. The phylogenetic tree, including 145 Bacteria, 18 Archaea, and 5 Eukarya, agrees with the "Bergey's Manual of Systematic Bacteriology" in a majority of lower taxa. In addition, we give some suggestions on higher taxa.3. Alignment-free methods are with less influence of computational amount and subjective factors compared with alignment methods. In this paper, alignment-free methods are used in the study of phylogenetic analysis of mitochondrion complete genome of mammalian. The result shows that it is more consistent with the traditional taxonomy by the use of FDOD. In addition, we find that FDOD do better than other alignment-free methods to unite members of the same order into one or few branches.