京都大学 / 0048 / 新制・課程博士 / 博士(農学) / 甲第21158号 / 農博第2284号 / 新制||農||1060(附属図書館) / 学位論文||H30||N5132(農学部図書室) / 京都大学大学院農学研究科地域環境科学専攻 / (主査)教授 田中 千尋, 教授 本田 与一, 教授 宮川 恒 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DGAM

A comparação de genomas é uma importante tarefa na qual a bioinformática pode ser aplicada, uma vez que ela permite a identificação de genes patogênicos, o que, por sua vez, pode auxiliar a combater ou a prevenir o surgimento de doenças. A partir da comparação de genomas, também é possível realizar a análise filogenética, que permite entender as relações evolutivas entre diferentes organismos. Em genomas de bactérias, essa análise geralmente é realizada com base no gene 16S rRNA. Entretanto, apesar de ser amplamente utilizado, filogenias com base nesse gene podem ter dificuldades para diferenciar organismos muito próximos evolutivamente. Essa importância da comparação de genomas e a necessidade de uma metodologia que permita distinguir organismos evolutivamente próximos na análise filogenética motivaram este trabalho, que teve como objetivo implementar ferramentas computacionais para identificar genes homólogos em genomas e, com base nesses genes, gerar filogenias e analisar se é possível distinguir os organismos evolutivamente próximos nessas filogenias. Para tanto, as ferramentas desenvolvidas para identificação de genes homólogos recebem resultados de alinhamentos e os filtram, de modo que dois genes são considerados homólogos se o alinhamento entre eles satisfizer os limiares definidos. Após a identificação das famílias de genes homólogos, tabelas são geradas com informações a respeito dos genes homólogos em cada genoma e, com base nessas tabelas, é possível gerar matrizes de distância e utilizar métodos de agrupamento hierárquico para a geração da filogenia ou realizar alinhamentos múltiplos com os genes identificados para posterior reconstrução filogenética. Além disso, também é possível representar os genes e famílias de genes homólogos por meio de um grafo, que pode auxiliar na escolha dos limiares para filtrar os alinhamentos. Para demonstrar e analisar a aplicabilidade das ferramentas desenvolvidas e das abordagens adotadas, experimentos foram realizados utilizando genomas de bactérias do gênero Xanthomonas, que contém um grande grupo de bactérias que causam doenças em plantas. Os resultados obtidos foram então comparados com filogenias de referência e com resultados de outros experimentos realizados. Essas comparações demonstraram que as famílias de genes homólogos podem ser úteis para distinguir genomas de organismos muito próximos evolutivamente, apesar de que essa abordagem apresentou dificuldades para separar os grupos de genomas mais distantes. Em contrapartida, na filogenia gerada a partir da região 16S rRNA, foi possível diferenciar esses organismos mais distantes, mas não foi possível distinguir os organismos muito próximos. Por fim, os experimentos realizados fornecem indícios de que as ferramentas desenvolvidas e as abordagens adotadas podem ser úteis para diferenciar genomas muito próximos evolutivamente de outros procariotos além das bactérias estudadas neste trabalho / Genome comparison is an important task on which bioinformatics can be used because it allows the identification of pathogen genes which can aid the combat of diseases and to avoid the emerging of new ones. Genome comparison also allows the phylogenetic analysis which provides the understanding of evolutional relations of different organisms. In bacterial genomes, this analysis is commonly based on 16S rRNA gene. Unfortunately, it can present some difficulties to distinguish closely related organisms. This importance of genome comparison and the necessity of a methodology to distinguish organisms that are closely related motivated this study, which aimed the development of computational tools to identify homologous genes in genomes, to use these genes to reconstruct phylogenies and to analyze if it is possible to distinguish closely related organisms on these phylogenies. To achieve this purpose, the developed tools to identify homologous genes receive the alignments results and filter it, such that two genes are homologous if their alignment satisfies the thresholds. After the identification of homologous gene families, the tools generates tables with information about the homologous genes presents in each genome and with these tables it is possible to create distance matrix to be used by hierarchical clustering methods to generate phylogenies or it is possible to perform multiple alignments with the identified genes to accomplish a phylogenetic reconstruction. Besides that, it is possible to represent the genes and homologous gene families in a graph, which can aid the choice of the thresholds to filter the alignments. To demonstrate and analyze the applicability of the developed tools and the approaches chosen in this study, experiments were performed using genomes of the bacterial genus Xanthomonas, which include a group of phytopathogenic bacteria. The results obtained were compared with reference phylogenies and with results of other experiments. These comparisons showed that homologous gene families can be used to differentiate closely related organisms, despite the fact that it presented difficulties to distinguish the groups of genomes that were evolutionarily far from each other. On the other hand, the phylogeny based on 16S rRNA region allows to distinguish the groups of genomes that were distant, but it was not possible to differentiate closely related organisms. As a conclusion, the experiments performed give pieces of evidence that the developed tools and the approaches adopted can be useful to distinguish genomes of closely related organisms of other prokaryotes besides the bacterias considered in this study

A comparação de genomas é uma importante tarefa na qual a bioinformática pode ser aplicada, uma vez que ela permite a identificação de genes patogênicos, o que, por sua vez, pode auxiliar a combater ou a prevenir o surgimento de doenças. A partir da comparação de genomas, também é possível realizar a análise filogenética, que permite entender as relações evolutivas entre diferentes organismos. Em genomas de bactérias, essa análise geralmente é realizada com base no gene 16S rRNA. Entretanto, apesar de ser amplamente utilizado, filogenias com base nesse gene podem ter dificuldades para diferenciar organismos muito próximos evolutivamente. Essa importância da comparação de genomas e a necessidade de uma metodologia que permita distinguir organismos evolutivamente próximos na análise filogenética motivaram este trabalho, que teve como objetivo implementar ferramentas computacionais para identificar genes homólogos em genomas e, com base nesses genes, gerar filogenias e analisar se é possível distinguir os organismos evolutivamente próximos nessas filogenias. Para tanto, as ferramentas desenvolvidas para identificação de genes homólogos recebem resultados de alinhamentos e os filtram, de modo que dois genes são considerados homólogos se o alinhamento entre eles satisfizer os limiares definidos. Após a identificação das famílias de genes homólogos, tabelas são geradas com informações a respeito dos genes homólogos em cada genoma e, com base nessas tabelas, é possível gerar matrizes de distância e utilizar métodos de agrupamento hierárquico para a geração da filogenia ou realizar alinhamentos múltiplos com os genes identificados para posterior reconstrução filogenética. Além disso, também é possível representar os genes e famílias de genes homólogos por meio de um grafo, que pode auxiliar na escolha dos limiares para filtrar os alinhamentos. Para demonstrar e analisar a aplicabilidade das ferramentas desenvolvidas e das abordagens adotadas, experimentos foram realizados utilizando genomas de bactérias do gênero Xanthomonas, que contém um grande grupo de bactérias que causam doenças em plantas. Os resultados obtidos foram então comparados com filogenias de referência e com resultados de outros experimentos realizados. Essas comparações demonstraram que as famílias de genes homólogos podem ser úteis para distinguir genomas de organismos muito próximos evolutivamente, apesar de que essa abordagem apresentou dificuldades para separar os grupos de genomas mais distantes. Em contrapartida, na filogenia gerada a partir da região 16S rRNA, foi possível diferenciar esses organismos mais distantes, mas não foi possível distinguir os organismos muito próximos. Por fim, os experimentos realizados fornecem indícios de que as ferramentas desenvolvidas e as abordagens adotadas podem ser úteis para diferenciar genomas muito próximos evolutivamente de outros procariotos além das bactérias estudadas neste trabalho / Genome comparison is an important task on which bioinformatics can be used because it allows the identification of pathogen genes which can aid the combat of diseases and to avoid the emerging of new ones. Genome comparison also allows the phylogenetic analysis which provides the understanding of evolutional relations of different organisms. In bacterial genomes, this analysis is commonly based on 16S rRNA gene. Unfortunately, it can present some difficulties to distinguish closely related organisms. This importance of genome comparison and the necessity of a methodology to distinguish organisms that are closely related motivated this study, which aimed the development of computational tools to identify homologous genes in genomes, to use these genes to reconstruct phylogenies and to analyze if it is possible to distinguish closely related organisms on these phylogenies. To achieve this purpose, the developed tools to identify homologous genes receive the alignments results and filter it, such that two genes are homologous if their alignment satisfies the thresholds. After the identification of homologous gene families, the tools generates tables with information about the homologous genes presents in each genome and with these tables it is possible to create distance matrix to be used by hierarchical clustering methods to generate phylogenies or it is possible to perform multiple alignments with the identified genes to accomplish a phylogenetic reconstruction. Besides that, it is possible to represent the genes and homologous gene families in a graph, which can aid the choice of the thresholds to filter the alignments. To demonstrate and analyze the applicability of the developed tools and the approaches chosen in this study, experiments were performed using genomes of the bacterial genus Xanthomonas, which include a group of phytopathogenic bacteria. The results obtained were compared with reference phylogenies and with results of other experiments. These comparisons showed that homologous gene families can be used to differentiate closely related organisms, despite the fact that it presented difficulties to distinguish the groups of genomes that were evolutionarily far from each other. On the other hand, the phylogeny based on 16S rRNA region allows to distinguish the groups of genomes that were distant, but it was not possible to differentiate closely related organisms. As a conclusion, the experiments performed give pieces of evidence that the developed tools and the approaches adopted can be useful to distinguish genomes of closely related organisms of other prokaryotes besides the bacterias considered in this study

No description available.

Yersinia pestis, a gram-negative bacterium, is the causative agent of plague. Y. pestis is a zoonotic pathogen that occasionally infects humans, and is endemic in the western United States. History gives evidence of three main plague pandemics. The first, originating in Egypt in 541AD, is known as the Justinian plague. The second, perhaps most well-known, is thought to have emerged in 1347AD in China, and is called the Black Death. The third, and current plague pandemic, also emerged in China in 1855. In 1899, Y. pestis was established in California, and the plague in other parts of America evolved from this initial introduction. In order to understand evolutionary patterns, we sequenced and analyzed 22 novel Y. pestis genomes from New Mexico. Performing a multiple genome alignment was the first step of our computational pipeline, after which evolutionary patterns were elucidated. Results from this study include predictions of four genes under negative selection pressure. Three of these genes were located on the Y. pestis chromosome, the fourth on the pCD1 virulence plasmid. This study also revealed 42 sites displaying statistically significant skew in the observed residue distribution when comparing sequences based on the year of isolation, and nine significant sites when comparing sequences based on the host species. Phylogenetic tree reconstruction showed a monophyletic pattern for sequences collected in the United States. Taken together, these analyses shed light on the evolutionary history of this pathogen in the southwestern US over a focused time range.

Cyanobakterien sind meist photoautotroph lebende Prokaryoten, welche nahezu alle Biotope der Welt besiedeln. Sie gehören zu den wichtigsten Produzenten der weltweiten Nahrungskette. Um sich auf den täglichen Wechsel von Tag und Nacht einzustellen, besitzen Cyanobakterien eine innere Uhr, bestehend aus den Proteinen KaiA, KaiB und KaiC, deren biochemische Interaktionen zu einem 24-stündigen Rhythmus von Phosphorylierung und Dephosphorylierung führen. Die circadiane Genexpression im Modellorganismus Synechocystis sp. PCC 6803 habe ich mittels drei verschiedener Zeitserienexperimente untersucht, wobei ich einen genauen Zeitplan der Genaktivierung in einer Tag-Nacht-Umgebung, aber keine selbsterhaltenden Rhythmen entdecken konnte. Allerdings beobachtete ich einen überaus starken Anstieg der ribosomalen RNA in der Dunkelheit. Aufgrund ihrer hohen Wachstumsraten und der geringen Anforderungen an die Umwelt bilden Cyanobakterien eine gute Grundlage für die nachhaltige Erzeugung von Biokraftstoffen, für einen industriellen Einsatz sind aber weitere Optimierung und ein verbessertes Verständnis des Metabolismus von Nöten. Hierfür habe ich die Orthologie von verschiedenen Cyanobakterien sowie die Konservierung von Genen und Stoffwechselwegen untersucht. Mit einer neu entwickelten Methode konnte ich gemeinsam vorkommende Gene identifizieren und zeigen, dass diese Gene häufig an einem gemeinsamen biologischen Prozess beteiligt sind, und damit bisher unbekannte Beziehungen aufdecken. Zusätzlich zu den diskutierten Modulen habe ich den SimilarityViewer entwickelt, ein grafisches Computerprogramm für die Identifizierung von gemeinsam vorkommenden Partnern für jedes beliebige Gen. Des Weiteren habe ich für alle Organismen automatische Rekonstruktionen des Stoffwechsels erstellt und konnte zeigen, dass diese die Synthese von gewünschten Stoffen gut vorhersagen, was hilfreich für zukünftige Forschung am Metabolismus von Cyanobakterien sein wird. / Cyanobacteria are photoautotrophic prokaryotes populating virtually all habitats on the surface of the earth. They are one of the prime producers for the global food chain. To cope with the daily alternation of light and darkness, cyanobacteria harbor a circadian clock consisting of the three proteins KaiA, KaiB, and KaiC, whose biochemical interactions result in a phosphorylation cycle with a period of approximately 24 hours. I conducted three time-series experiments in the model organism Synechocystis sp. PCC 6803, which revealed a tight diurnal schedule of gene activation. However, I could not identify any self-sustained oscillations. On the contrary, I observed strong diurnal accumulation of ribosomal RNAs during dark periods, which challenges common assumptions on the amount of ribosomal RNAs. Due to their high growth rates and low demand on their environment, cyanobacteria emerged as a viable option for sustainable production of biofuels. For an industrialized production, however, optimization of growth and comprehensive knowledge of the cyanobacterial metabolism is inevitable. To address this issue, I analyzed the orthology of multiple cyanobacteria and studied the conservation of genes and metabolic pathways. Systematic analysis of genes shared by similar subsets of organisms indicates high rates of functional relationship in such co-occurring genes. I designed a novel approach to identify modules of co-occurring genes, which exhibit a high degree of functional coherence and reveal unknown functional relationships between genes. Complementing the precomputed modules, I developed the SimilarityViewer, a graphical toolbox that facilitates further analysis of co-occurrence with respect to specific cyanobacterial genes of interest. Simulations of automatically generated metabolic reconstructions revealed the biosynthetic capacities of individual cyanobacterial strains, which will assist future research addressing metabolic engineering of cyanobacteria.

No description available.