DIMEDR

DIMEDR harmonizes incongruent LC-MS metabolomics datasets of identical biological sample types by creating a unified framework that enhances spectral feature similarity across disparate datasets to enable comparative analyses.

Key Features:

• Dataset harmonization: Harmonizes incongruent LC-MS metabolomics datasets derived from identical biological sample types to improve inter-dataset comparability.
• Primary-dataset templating: Selects and postprocesses a "primary" dataset using peak identification, alignment, and grouping to serve as a progenitor template for integration.
• Universal retention time correction: Performs universal retention time correction and comparison leveraging ubiquitous features from the primary dataset.
• Endogenous internal standards: Uses ubiquitous spectral features identified in the primary dataset as endogenous internal standards during integration.
• Spectral feature similarity enhancement: Creates a unified analytical framework that enhances spectral feature similarity across disparate datasets.
• Demonstrated integration: Validated by unifying two human and two mouse urine LC-MS metabolomics datasets collected over four years from unrelated studies.

Scientific Applications:

• Cross-study comparative analyses: Enables meaningful cross-study comparisons of LC-MS metabolomics data that are otherwise incompatible.
• Multi-year dataset integration: Facilitates integration of datasets collected across multi-year periods and disparate studies, as shown with human and mouse urine datasets.
• Comparative metabolic investigations: Supports comparative analyses of metabolic processes across biological contexts by reassembling disparate LC-MS datasets into a cohesive structure.

Methodology:

Select and postprocess a "primary" dataset via peak identification, alignment, and grouping; identify ubiquitous features to act as endogenous internal standards; apply universal retention time correction and comparison; reassemble and integrate disparate LC-MS datasets into a unified framework to enhance spectral feature similarity.