MetAssimulo

MetAssimulo simulates ^1H-NMR spectra of complex metabolic mixtures in MATLAB to support development and benchmarking of metabolomics and metabonomics data-analysis methods.

Key Features:

• Data Source Integration: Integrates a metabolite standard spectral database with concentration information supplied by the user or automatically constructed from the Human Metabolome Database.
• Simulation of Complex Profiles: Generates realistic metabolic profiles containing numerous metabolites and constructs two groups ('case' and 'control') from specified means and standard deviations of metabolite concentrations.
• Noise Simulation: Adds spectral noise with a realistic autocorrelation structure as specified by the user.
• Correlation Analysis: Simulates intra- and inter-metabolite correlations present in metabolic data.
• Matrix Effect Simulation: Models shifts in NMR peak positions due to matrix effects such as pH differences.

Scientific Applications:

• Algorithm development and benchmarking: Provides simulated ^1H-NMR metabolic profiles to develop and test data analysis and statistical methods without requiring experimental sample collection.
• Method evaluation under realistic variability: Enables evaluation of the robustness of chemometric and statistical algorithms to noise, correlation structures, and matrix-induced peak shifts.
• Study of genetic and environmental effects on metabolism: Supports simulation studies of genetic and environmental interactions affecting metabolite concentrations and spectral patterns.

Methodology:

MATLAB-based algorithm integrates metabolite standard spectral data with concentration information (user-provided or derived from the Human Metabolome Database), constructs 'case' and 'control' groups from specified concentration means and standard deviations, imposes intra- and inter-metabolite correlation structures, adds autocorrelated spectral noise, and models matrix-induced NMR peak shifts such as those caused by pH differences.