Identifying bacterial strains is important not only for human health research but also for various environmental and industrial applications. The plethora of available shotgun metagenomic datasets provide an unprecedented opportunity to discover novel bacterial strains in different environmental niches. In this thesis, we evaluated the existing novel-strain-based tools and showed that their performance is still suboptimal. Due to the difficulty in distinguishing strains of similar abundance by available tools, the SMS (strains in multiple samples) tool was developed to de novo identify bacterial strains in multiple shotgun metagenomic samples. We showed that SMS distinguishes strains of similar abundance well and outperforms other novel-strain-based tools on both simulated and experimental datasets. Applying the SMS tool to the Atopic Dermatitis (AD) samples, we discovered novel strains of Staphylococcus aureus and Staphylococcus epidermidis, and diversity that could not be observed when the analysis was based on current known strain databases. Compared with the previously identified known strains in the same samples, the predicted novel strains showed a better likelihood of being present in the samples than the known strains. Annotation and functional pathway analysis of these AD-related novel strains revealed their relation to coding sequences and functional pathways that have been known to contribute to the ability that bacteria become pathogenic and carry out infection, suggesting their potential roles in AD progression.
If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu
Doctor of Philosophy (Ph.D.)
College of Medicine
Burnett School of Biomedical Sciences
Length of Campus-only Access
Doctoral Dissertation (Open Access)
Ventolero, Minerva Fatimae, "Computational Strain Analysis in Microbiome Datasets with Application to Skin Disease" (2022). Electronic Theses and Dissertations, 2020-. 1687.