Bacterial Genome In The Genetic Makeup Of Nucleic Acid And The Comparative Analysis And Theoretical Predictions

With the fast development of large-scale sequencing, large amounts of data are available in public sequence depository. Then how to mine useful biological meanings is becoming a pressing issue. In this thesis, we analyzed genome compositions of all currently available sequenced bacterial genomes in GenBank by bioinformatics methods. The whole process and results can be divided into the following four parts:Firstly, we investigated the relationships between GC content and overall codon usage of bacteria genomes and a similar codon usage pattern via GC content was found existed in all bacteria genomes. Meanwhile, we studied the relations of genome GC content and chromosome length, and a gamma distribution was found of GC content per chromosome length. In other words, there is an overall consistency existing in all bacterial genome compositions.Secondly, in a single chromosome, there also exists non-consistency. The properties of ununiformed base distribution on chromosome, plus the connection between structure and function, can be used in identifying genomic islands in bacteria genomes. By using segmentation algorithm, combined with Z curve method, more than50genomic islands are found in nine sequenced Bacillus Cereus.Thirdly, even in the same strain of bacteria, the composition on different chromosomes of multiple chromosome strain is significant different. A unique phenomenon is found in B. cenocepacia AU1054that the shortest chromosome of AU1054has much more essential genes and tRNA genes than the corresponding chromosomes in the other10strains. The present work may contribute to the understanding of how the secondary chromosomes of multipartite bacterial genomes originate and evolve.At final stage, after analyzing similarities and differences of genomic compositions at different levels, we try to take advantage of the composition parameters, combined with machine learning algorithms, to predict essential genes in bacteria genomes. In particular, E. coli and Mycoplasma were studied separately. Then all currently available essential genes in16species were combined and trained, and then a predicted model was obtained.In a word, we dig into the composition of bacteria genome, the similarities and differences of the genomic compositions were showed in the analysis. The basic analysis on different levels of genome shows something hidden deeply from the composition.