Joshua M. Mitchell

The paper analyzes metabolomics 'dark matter' using a detectable khipu model, showing most abundant features have identifiable ion patterns, while many remain unknown.

The paper constructs a consensus serum metabolome from over 100,000 mass spectrometry acquisitions, revealing gaps in current databases and improving LC-MS metabolomics annotation.

The paper introduces a Python pipeline and data structures for metabolomics data processing, annotation, and standardization, demonstrating its utility with large-scale datasets.

The paper introduces a scan-centric method for characterizing peaks in mass spectrometry data, improving the accuracy of isotope abundance analysis.

The paper introduces a graph coloring method to generate atom identifiers for compounds, enabling compound harmonization across metabolic databases and improving metabolic network integration.

The paper develops a machine learning method to classify molecular formula assignments into lipid categories using SMIRFE data, achieving high accuracy and precision.

The paper identifies and addresses high peak density artifacts in Fourier transform mass spectra, which can mislead data analysis by creating false peaks.

The paper develops a Bayesian graphical model to infer metabolite interactomes using molecular structure information and applies it to coronary artery disease data.