Analysis and correction of crosstalk effects in pathway analysis

Area of Science:

• Systems Biology
• Bioinformatics
• Genomics

Background:

• Pathway analysis is key to understanding biological conditions.
• Current methods rely on P-values to assess pathway significance.
• P-values were assumed to be independent.

Purpose of the Study:

• To investigate the phenomenon of pathway crosstalk and its impact on P-values.
• To develop a method for detecting, quantifying, and correcting pathway crosstalk.
• To identify independent functional modules for improved biological interpretation.

Main Methods:

• Analysis of three major pathway analysis categories: enrichment analysis, functional class scoring, and topology-based methods.
• Development of a novel technique to quantify and correct for crosstalk effects.
• Validation of the approach on real-world experimental data from multiple phenotypes and species.

Main Results:

• Pathway crosstalk significantly influences P-values across all major analysis methods.
• Crosstalk can lead to biologically meaningless pathways appearing significant.
• Crosstalk correction can reveal significant pathways that were previously non-significant.
• The developed method successfully identifies independent functional modules.

Conclusions:

• Pathway P-values are interdependent due to crosstalk, challenging existing analysis methods.
• A novel technique effectively quantifies and corrects for pathway crosstalk.
• This approach enhances the accuracy of pathway analysis and biological interpretation.
• The method aids in refining signaling pathway definitions for specific phenotypes.