Can We Assume the Gene Expression Profile as a Proxy for Signaling Network Activity?

Area of Science:

• Systems biology
• Molecular biology
• Bioinformatics

Background:

• Gene expression profiling is common in systems biology, often assuming transcript and protein correlation.
• The relationship between gene product interactions (activation/inhibition) and expression profiles is understudied.
• Current methods often focus on differentially expressed genes, overlooking effects on signaling pathway activity.

Purpose of the Study:

• To investigate if significant gene expression changes necessarily cause dysregulated signaling pathways.
• To evaluate the coherency between gene expression levels and causal relationships in signaling networks.
• To determine if transcript-level data is sufficient for inferring pathway perturbations.

Main Methods:

• Extracted gene expression data and gene pair relationships from four comprehensive databases.
• Assessed the sign consistency between expression levels and causal interactions of gene pairs.
• Compared signaling network coherency against random gene pairs.

Main Results:

• The signaling network exhibits incoherency and inconsistency with observed expression profiles.
• Random gene pairs showed a different pattern than actual signaling network interactions.
• A significant portion of gene expression changes did not align with predicted pathway dysregulation.

Conclusions:

• Gene expression data alone is insufficient for inferring dysregulated signaling pathways.
• The type of gene product relationship (activation/inhibition) must be considered alongside expression data.
• Relying solely on differentially expressed genes for pathway inference is limited and may yield inaccurate results.