iMTBGO: An Algorithm for Integrating Metabolic Networks with Transcriptomes Based on Gene Ontology Analysis.

Area of Science:

• Systems Biology
• Metabolic Engineering
• Computational Biology

Background:

• Constraint-based metabolic network models are vital for predicting phenotypes and designing metabolic engineering strategies.
• Integrating 'omics' data, particularly transcriptomics, with these models aims to enhance prediction accuracy.
• Existing flux balance analysis (FBA)-based algorithms often integrate gene expression data.

Purpose of the Study:

• To develop a novel method for predicting metabolic flux distributions by integrating gene expression data.
• To improve the accuracy of metabolic flux predictions compared to existing methods.
• To leverage Gene Ontology (GO) term structure for normalizing gene expression data.

Main Methods:

• Mapped enzyme kinetic (Kcat) values to GO term hierarchy, observing higher similarity within the same GO terms.
• Developed the iMTBGO method, which constrains reaction boundaries using gene expression ratios normalized by marker genes within the same GO term.
• Applied iMTBGO to published data and compared its performance against other gene expression-informed metabolic flux analysis methods.

Main Results:

• Observed higher similarity in Kcat values for reactions associated with the same GO term.
• iMTBGO demonstrated smaller prediction errors for growth rates and central metabolic fluxes compared to previously published methods.
• The method effectively utilizes normalized gene expression data for improved metabolic flux prediction.

Conclusions:

• The iMTBGO method provides more precise metabolic flux predictions by incorporating gene expression values normalized via GO.
• This approach acknowledges that enzyme activity, influenced by gene expression, is crucial for determining reaction rates.
• The findings suggest a more accurate way to integrate transcriptomic data into metabolic models for biological insights and applications.