Heterogeneous Molecular Signatures in Staphylococcus aureus Infection Assessed by Multimodal Imaging

Abstract

Spatially targeted mass spectrometry approaches provide label-free characterization of tens to thousands of chemical species within a single experiment but often microscopy is needed to contextualize this molecular information. To do so, integrative computational methods that address the challenges of dimensionality, scaling, distribution, and multimodal visualization are required. This dissertation addresses these challenges through the development of new computational workflows for discovering relationships among measured molecular species and specific tissue features from multimodal mass spectrometry and microscopy data to characterize proteomic and lipidomic species within a Staphylococcus aureus infection. The development of these new methods is of great importance to the field of molecular imaging, leveraging powerful spatially aware methods alongside robust computational techniques to determine localized molecular changes and provide a systems biology-level summary of chemical changes. The application of these new computational methods enables a superior approach to analyzing and visualizing multimodal imaging data acquired from the staphylococcal host-pathogen interface.