Identification of perturbed signaling pathways from gene expression data using information divergence

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Abnormal signaling pathway regulation is implicated in various diseases.
• Existing methods often focus on individual gene expression or correlations within pathways, not overall pathway activity.

Purpose of the Study:

• To develop a new analysis model, SPAID (Signaling Pathway Analysis based on Information Divergence), for detecting overall changes in signaling pathways.
• To provide a more comprehensive framework for pathway analysis in different biological contexts.

Main Methods:

• SPAID utilizes information divergence to compare conditions by assessing differential probability distributions of pathway regulation capacity.
• The model was evaluated against classical algorithms using diverse microarray gene expression datasets.

Main Results:

• SPAID demonstrated higher repeatability and better overall performance compared to existing methods.
• The analysis extracted more comprehensive information regarding underlying biological processes.
• SPAID showed improved universality across different datasets.

Conclusions:

• Information divergence offers a novel perspective for measuring pathway differences holistically.
• SPAID provides a complementary analysis framework for understanding signaling pathway alterations in disease.
• The approach enhances the detection of functional pathway changes beyond individual gene metrics.