Mathias Alexander Witte Paz
The global rise in antibiotic resistance, coupled with the declining discovery of new compounds, has created a crisis that demands innovative strategies for microbial control. Identifying such novel strategies requires a mechanistic understanding of bacterial molecular pathways. The central dogma of molecular biology, expanded by newer findings on the flow of biological information, provides a fundamental roadmap for prokaryotic characterization based on high-throughput experiments and computational approaches. However, capturing this complexity requires integrative and reproducible computational frameworks that enable robust cross-layer analyzes. This thesis addressed this challenge by developing and applying computational methods that support integrative and visual analyses across the central dogma, with focus on microbial characterization for their control. To enhance the interactive characterization of prokaryotic genomes, the visual analytics tool Evidente is introduced, with the goal of bridging the gap between the exploration of single nucleotide polymorphisms (SNP) and the evolutionary context. Unlike traditional visualization tools, Evidente classifies SNPs based on clade-specificity and enables their visualization with metadata, as well as linking them to a functional context. By applying it to bacterial pathogens, this approach demonstrated how genome-scale variation can be interpreted in an evolutionary context to generate functional and phenotypic hypotheses. Recognizing that genomic data alone are insufficient to explain phenotypic diversity, this thesis integrated transcriptomic data through two complementary web-applications: TSSpredator-Web and TSS-Captur. They address the characterization of the transcriptome's architecture and link the genomic with the transcriptomic layer of the central dogma. TSSpredator-Web extends the tool TSSpredator to identify and classify transcription start sites (TSS), and allows exploration of genome-wide TSS maps together with genomic data. Based on TSS maps, the Nextflow-based pipeline TSS-Captur characterizes transcripts starting from unclassified TSS via computational methods for sequence classification, termination site prediction, and analyses of secondary structure and promoter regions. Together, both approaches showed how transcriptomic data can be used for annotation refinement and transcript discovery, bridging the genomic and transcriptomic layer. Lastly, the power of integrative analyses is demonstrated through two systematic studies of bacterial metallophore systems. The first project focused on the characterization of known metallophore mechanisms across the Staphylococcal genus using computational analyses based on genomic data, representing the first comprehensive genus-wide analysis of these systems. Expanding on this topic, the second study investigates how nasal Corynebacterium species exploit metallophores synthesized by Staphylococcus aureus, suggesting a novel way of controlling pathogens. This was achieved by identifying structural homologs of lipoproteins through the development of the Nextflow-based pipeline PRESERVE, and contextualizing the findings with transcriptomic data. By analyzing the putative promoter regions, we gained insight into how these mechanisms are regulated in Corynebacteria. These studies illustrate how the integration of multiple layers of biological data can shed light into the interactions between species, providing knowledge that can be translated into interference of colonization strategies, and be exploited for microbial control. In summary, this thesis describes a methodological framework for integrative computational analyses across the central dogma. By combining reproducible approaches with interactive visual exploration, it illustrates how integrating data from different biological layers can help to generate insights for prokaryotic characterizations. These insights can be used to establish innovative interventions for microbial control, offering a computational pathway to address the escalating challenges of antibiotic resistance.