No description available.

<p>Microbial metagenomics is the exploration of taxonomical diversity of microbial communities in environmental habitats using large, exhaustive DNA sequence datasets. However, due to inherent limitations of sequencing technology and the complexity of environmental genomes, current analytical approaches do not reveal the existence of all microbes that may be present. In this study, a new classification approach is proposed based upon unique proteins that are specific for different clades of Alphaproteobacteria to predict the presence and absence of species from these groups of bacteria in published metagenomic datasets. In this work, 264 previously–identified, published conserved signature proteins (CSPs) characteristic of individual taxonomic clades of Alphaproteobacteria are used as probes to detect the presence of bacteria in metagenomic datasets. Although public genome sequence information has increased manifold since these CSPs were initially identified 6 years ago, results indicate that nearly all of these CSPs (259 of 265) are specific for their previously characterized clades. Furthermore, they are confirmed to be present in the newly–identified and sequenced members of these clades. In view of their specificity and predictive ability in different monophyletic clades of Alphaproteobacteria, the sequences of these CSPs provide reliable probes to determine the presence or absence of these Alphaproteobacteria in metagenomic datasets. In this work, CSPs are used to determine the presence of Alphaproteobacteria diversity in 10 published metagenomic datasets (bioreactor, compost, wastewater, activated sludge, groundwater, freshwater sediment, microbial mat, marine, hydrothermal vent and whale fall metagenomes), which cover diverse environment and ecosystems. It is indicated that the BLAST searches with these CSPs can be used to efficiently identify Alphaproteobacteria species in these metagenome dataset and substantial differences can be determined in the distribution and relative abundance of different Alphaproteobacteria species in the tested metagenome datasets. Thus the CSPs, which are specific for different microbial taxa, provide novel and powerful means for identification of microbes and for their taxonomic profiling in metagenomic datasets.</p> / Master of Science (MSc)

La miniaturisation des systèmes de piégeage optique permet la manipulation et l’analyse d’objets de taille nano et micrométrique. Ces objets peuvent être inertes (billes de silice ou de polystyrène, nanotubes de carbones) ou biologiques (virus, bactéries, cellules). Les dispositifs de piégeage intégrés sur puces présentent l’avantage de manipuler des objets uniques de manière réversible et à très faible puissance. Cela en fait des systèmes très adaptés à l’étude des objets biologiques, souvent plus fragiles et traditionnellement étudiés à l’échelle de la population dans les méthodes de microbiologie. Ces travaux de thèse portent sur l’étude de différentes bactéries piégées sur cristaux photoniques 1D et 2D. Nous y démontrons tout d’abord le piégeage de sept espèces bactériennes : E. coli, B. subtilis, S. epidermidis, Y. ruckeri, N. sicca, P. putida et L. innocua. . Nous y décrivons les méthodes de caractérisation spatiale et temporelle développées pour extraire de l’information de ce piégeage sur la taille, la forme, le déplacement et la structure membranaire des bactéries. Un dispositif de piégeage à deux lasers a également été implémenté pour permettre l’analyse fine de l’état d’une bactérie E. coli soumise à un stress thermique. Ces résultats s’inscrivent dans une problématique de diagnostic bactérien, très sensible depuis quelques années avec l’augmentation de la résistance des bactéries aux antibiotiques. Si aucun changement n’intervient, il est prédit que les infections bactériennes constitueront la première cause de mortalité des pays développés d’ici 2050. Il est donc nécessaire d’élaborer de nouveaux outils de diagnostic plus rapides et plus accessibles afin de limiter la distribution abusive d’antibiotiques qui entraine ce phénomène de résistance Nous proposons également en dernière partie de ce manuscrit des solutions pour intégrer notre dispositif de piégeage en vue d’ouvrir la voie à de nouvelles applications dans des environnements pathogènes. / The miniaturisation of optical trapping systems allows the manipulation and analysis of nano and micron sized objects. These objects can be inert (silica or polystyrene beads, carbon nanotubes) or biological (viruses, bacteria, cells). Integrated trapping devices on chips have the advantage of handling single objects reversibly and at very low power. This makes them very suitable for the study of biological objects, often more fragile and traditionally studied at the population level in microbiology methods. This work deals with the study of different bacteria trapped on 1D and 2D photonic crystals. We first demonstrate the trapping of seven bacterial species: E. coli, B. subtilis, S. epidermidis, Y. ruckeri, N. sicca, P. putida and L. innocua. . We describe a spatial and temporal characterisation methods developed to extract information from this trapping on the size, shape, motility and membrane structure of bacteria. A trapping device with two lasers has also been implemented to allow the fine analysis of the state of an E. coli bacterium subjected to heat stress. These results falls within the issue of bacterial diagnosis, very sensitive in recent years with the increase of the resistance of bacteria to antibiotics. If no change occurs, it is predicted that bacterial infections will be the main cause of death in developed countries by 2050. It is therefore necessary to develop new, faster and more accessible diagnostic tools to limit the large distribution of antibiotics leading to this phenomenon of antibioresistance. We also propose in the last part of this manuscript solutions to integrate our trapping device in order to pave the way for new applications in pathogenic environments.