In The Shadows Of The Metabolome: Unknown Spectral Features In Soil, Scent, And Disease

Untargeted metabolomics offers a powerful approach for analyzing small molecules in complex biological and environmental samples. This dissertation applies high resolution mass spectrometry to three distinct systems—soil chemistry, forensic residual odor detection, and clinical metabolomics—to explore how spectral features can reveal meaningful chemical differences, even without full structural identification. The first study examined how land management influences soil metabolomes. Despite limited metabolite annotations, spectral data revealed distinct chemical profiles across forested, cultivated, and pasture soils. These patterns suggest that land use alters nutrient cycling and elemental composition in reproducible ways. The second study tested whether cadaver detection dogs can respond to residual scent left on materials exposed briefly to decomposing remains. Untargeted GC HRMS analysis showed clear chemical signatures, including thiols and sulfur containing volatiles, supporting canine alerts to trace level residues even after direct odor emission had ceased. The final study investigated blood serum from individuals with and without neurodegenerative disease. Metabolic differences between groups were observed, highlighting potential chemical markers linked to disease processes. Together, these studies demonstrate the utility of untargeted metabolomics for detecting subtle but meaningful biochemical variation across diverse applications. By focusing on patterns within unknown spectral features, this work shows how mass spectrometry can advance ecological, forensic, and clinical understanding without requiring complete compound identification.