Computational approaches for metatranscriptomic profiling in translational medicine and pulmonary diseases

Use of total RNA-seq in host and microbiome analysis allows for multi-omic interrogation of microbial profiles, assessment of their function and their interaction with host immune and metabolic pathways. This type of analysis calls for novel computational techniques. However, existing tools for analyzing microbial multi-omic data are lacking, as they typically address a single data type. For example, there are many available tools for the characterization of microbial communities, but these are unable to investigate microbial-host interactions. To address this need, I developed a novel computational pipeline that integrates existing methods for microbial and host expression profiling. This pipeline provides insight into possible personalized medical interventions in translational medicine. This dissertation utilized — transcriptomics and metatranscriptomics to interrogate: 1) host-microbial interactions in people with indeterminate pulmonary nodules, 2) the role of Human Endogenous Retroviruses in the early onset of ageing observed in virologically suppressed HIV positive individuals, and finally 3) to characterize humoral responses to SARS-CoV-2 peptides in Covid-19 patients. Specifically, to address the host-microbial interactions in people with indeterminate pulmonary nodules, I addressed sources of batch effects in the data, and I utilized statistical approaches to identify differentially abundant microbes in current and former smokers and malignant and benign samples. Lastly, I linked abundant microbes in both datasets to human pathways and tested for their strength of association. This approach aided in providing insight into the possible functional profile of these microbes and their role in lung cancer. Furthermore, I investigated the role of Human Endogenous Retroviruses in the early onset of ageing observed in virologically suppressed HIV positive individuals. In this project, I utilized Telescope software to generate HERVs counts. Differential analyses were then performed to identify differentially expressed HERVs in PLHIV. Using the computational pipeline that was developed for muti-omic analyses, the association of differentially expressed HERVs with pathways involved in inflammageing and inflammatory markers was then investigated. Taken together, this work identified HERVS that could act as therapeutic and diagnostic in the HIV setting. Lastly, for the third project, I sought to characterize IgG and IgM humoral responses to SARS-CoV-2 at the epitope level, where discriminating epitopes for disease severity were identified. I also investigated epitopes that were conserved between SARS-CoV-2 virus and other Human coronaviruses, allowing the investigation of associations with less severe disease outcomes. These epitopes could serve as discriminative markers for COVID-19 disease severity.2026-01-11T00:00:00