SIRM

SIRM performs moiety-level modeling and model selection using stable isotope-resolved metabolomics (SIRM) isotopologue and isotopomer data from mass spectrometry (MS) to support metabolic flux analysis and metabolic modeling.

Key Features:

• Moiety Model Comparison and Selection: Compares multiple moiety models against experimental metabolomics data provided in JSON to select the best-supported candidate models.
• Model Parameter Optimization: Performs parameter estimation and analysis of optimization results to identify best-fitting model parameters.
• Multi-Tracer Dataset Analysis: Supports analysis of datasets derived from multi-tracer stable isotope experiments.
• Robustness and Flexibility: Validated on time-series MS isotopologue datasets for uridine diphosphate N-acetyl-D-glucosamine (UDP-GlcNAc) from different MS platforms and capable of selecting optimal models from pools of over 40 candidates.
• Error Mitigation and Optimization: Enables selection of objective functions and optimization criteria to minimize side effects of uncertainty and reduce over-optimization during model fitting.
• Combination of Datasets: Integrates SIRM datasets, including public and newly acquired data, to improve model selection and mitigate overfitting.

Scientific Applications:

• Metabolic Flux Analysis: Supports inference of metabolic fluxes from isotopologue and isotopomer MS data.
• Metabolic Pathway Analysis: Aids elucidation and discrimination of alternative pathway models through moiety-level model selection.
• Comparative Metabolomics Studies: Facilitates comparison of moiety models across conditions or organisms to investigate metabolic variation.
• Data Integration and Curation: Promotes use of curated SIRM datasets to improve robustness and reproducibility of metabolic models.

Methodology:

Uses JSONized metabolomics datasets to compare time-series isotopologue profiles against candidate moiety models, applies optimization methods, criteria, and objective functions for parameter estimation and model selection, performs error analysis and mitigation via objective-function choice, and is implemented in the moiety_modeling Python package.