G-Aligner

G-Aligner applies a graph-based method to align features across untargeted liquid chromatography–mass spectrometry (LC-MS) runs, performing retention time correction and comprehensive feature matching to support accurate metabolomic profiling.

Key Features:

• Graph-Based Approach: Represents features and potential correspondences as nodes and edges in a multipartite graph to model complex inter-run relationships.
• Multidimensional Assignment Problem (MAP): Frames feature matching as an unbalanced multidimensional assignment problem and uses three combinatorial optimization algorithms to identify optimal matches.
• Comprehensive Feature Correspondence: Considers correspondences among all runs rather than focusing solely on retention time correction to reduce mismatches.
• Performance Superiority: Outperformed existing methods in OpenMS, MZmine2, and XCMS on three public metabolomics benchmark datasets, achieving up to a 9.8% increase in accurately aligned features and a 26.6% increase in analytes.
• Integration with Existing Tools: Integrates into existing analysis workflows to refine self-extracted features from other LC-MS processing software.

Scientific Applications:

• Untargeted Metabolomics Feature Alignment: Improves alignment accuracy across multiple LC-MS runs to support reliable comparative metabolomic studies.
• Metabolite Identification and Quantification: Enhances matching accuracy to aid identification and quantification of metabolites for metabolic pathway analysis and biomarker discovery.

Methodology:

Constructs a multipartite graph with features as nodes and potential correspondences as edges, then solves the unbalanced multidimensional assignment problem using three combinatorial optimization algorithms.