ROIMCR

ROIMCR resolves LC-MS metabolomic data into pure chemical components by identifying regions of interest (ROIs) in the m/z domain, compressing the data into a feature matrix, and applying Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) to support untargeted metabolomics analyses.

Key Features:

• Data filtering and compression: Identifies ROIs within the mass-to-charge ratio (m/z) domain and compresses large LC-MS datasets into a matrix of features while preserving spectral accuracy.
• MCR-ALS decomposition: Applies Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) to resolve compressed data into contributing pure components.
• No chromatographic alignment or peak shaping required: Resolves components without prior chromatographic alignment or peak shaping, avoiding potential distortions from those preprocessing steps.
• ROI-focused compression versus binning/windowing: Uses an ROI-based compression strategy that maintains spectral integrity and surpasses conventional binning and windowing approaches in retaining relevant information.

Scientific Applications:

• Untargeted metabolomics: Enables component resolution and feature extraction for complex untargeted LC-MS metabolomic studies.
• Lipidomics: Applicable to LC-MS lipidomics datasets for resolving lipid-related components.
• Large-scale LC-MS studies: Suited for cross-disciplinary and large-volume LC-MS investigations requiring efficient data reduction and component deconvolution.

Methodology:

The approach identifies ROI features within the m/z domain, compresses the data into a feature matrix, and applies MCR-ALS to resolve pure components without chromatographic alignment or peak shaping; the method is implemented in MATLAB.