Analysis Of Gene And Genome Structures In Prokaryotes

With the implementation of a large scale of genome sequence project, a great amount of genome sequences of prokaryotic and eukaryotic organisms have been accomplished. It is the most pressing task for analyzing the sequences at gene and genome levels. Prokaryotic genomes are the preferred objection to study the genetic characteristics of the biological molecular because of their small genomes and simple structure. In this thesis, we are devoted to discussing the structure features of prokaryotic genes and genomes using the method of nonlinear mapping (NLM). In addition, prediction of replication origins in some archaebacteria genomes is briefly introduced.In chapter I , bioinformatics are first introduced, and then new achievements referred to the work in my thesis are presented.In chapter II, the biological knowledge related to study in the thesis is explained, involving the central dogma, the semi-conservative replication and the semi-discontinuous replication, the basic features of codons, and so on.In chapter III, the multivariate analysis is described. The multivariate method is an effective tool to analyze the data in the multidimensional space. In this paper, the emphasis is placed on the technique for reducing the dimensions. The principal analysis, correspondence analysis and nonlinear mapping are described in detail.In chapter IV, the codon usage preferences in prokaryotic organism are analyzed by nonlinear mapping method. The results show that codon usage between highly expressed genes such as ribosomal protein genes and the others is quite different. It is reason that the highly expressed genes prefer to use codons corresponding to the most abundant tRNAs in the cell. So the asymmetrical replication and the translational selection determine the pattern of synonymouscodon usage in prokaryotic organisms.Besides the work, an approach on predicting the replication origin in archaebacteria genomes is proposed based on the asymmetry of base contents on two sides of replication origin in prokaryotic genomes. The approach can recognize replication origins well.