Bacterial Plant Pathogen Identification using Genomics and Metagenomics

The timely identification of pathogens responsible for disease outbreaks is crucial for implementing effective control measures and minimizing the spread of infectious diseases. Conventional methods of identification are limited to specific pathogen species because they require prior knowledge and pure cultures of the pathogen. Therefore, these methods cannot detect new pathogens responsible for newly emerging diseases. Computational methods that rely on sequencing data have the potential to overcome these limitations. However, the diverse phenotypes among microbial species and strains within the same species pose a challenge in accurately identifying the specific pathogen responsible for the disease. This dissertation highlights the importance of strain-level detection for the identification and characterization of pathogens by employing computational methods that rely on genomic and metagenomic sequencing data. To realize that computational goal, a comparison of different tools, currently used for metagenome classification, was done to illustrate effective detection of bacterial pathogens. To develop computational methods for characterization, genomes of the plant pathogen Ralstonia solanacearum were studied to understand the basis of virulence at cool temperatures. Finally, a new tool was developed that combines accurate detection and characterization at the strain level, through the use of taxonomic databases constructed using genome similarity thresholds. This dissertation work is a contribution to the development of improved approaches to detect and contain disease outbreaks in plants with possible applications in human and animal diseases as well.