Selective of informative metabolites using random forests based on model population analysis.

Area of Science:

• Metabolomics
• Biomarker Discovery
• Computational Biology

Background:

• Metabolomics studies aim to identify disease biomarkers for diagnosis and pathology.
• Extracting relevant information from complex 'omics' data requires advanced feature selection methods.

Purpose of the Study:

• To propose a novel and robust method for selecting informative metabolites from metabolomic datasets.
• To classify metabolites into informative, non-informative, and interfering categories based on their contribution to classification accuracy.

Main Methods:

• A new selective method combining random forests (RF) with model population analysis (MPA).
• Classification of metabolites based on their contribution to classification accuracy.
• T-test analysis to validate the significance of selected metabolites between healthy and diseased groups.

Main Results:

• The proposed MPA-RF method successfully selected informative metabolites from three metabolomic datasets.
• Selected metabolites showed statistically significant differences (P < 0.05) between healthy and diseased groups.
• Identified informative metabolites correlated with the clinical outcome under investigation.

Conclusions:

• The MPA-RF method is effective for identifying clinically relevant biomarkers in metabolomics.
• This approach enhances the discovery of diagnostic and prognostic indicators for diseases.
• The developed method provides a robust tool for feature selection in complex biological data.