PIGNON: a protein-protein interaction-guided functional enrichment analysis for quantitative proteomics

Area of Science:

• Proteomics
• Systems Biology
• Bioinformatics

Background:

• Quantitative proteomics identifies differentially expressed proteins.
• Functional enrichment analysis (FEA) uses these proteins to infer biological process dysregulation.
• Standard FEA relies on arbitrary significance thresholds, potentially missing subtle but coordinated changes in functional annotations.

Purpose of the Study:

• To introduce PIGNON, a novel graph theory-based method for detecting differentially expressed functional annotations.
• To overcome limitations of standard FEA by considering protein-protein interaction networks.
• To provide a more comprehensive characterization of quantitative proteomics data.

Main Methods:

• PIGNON maps protein differential expression levels onto a protein-protein interaction network.
• It measures the network clustering of proteins within specific functional annotations.
• A Monte-Carlo sampling approach assesses the significance of this clustering in an expression-weighted network.

Main Results:

• PIGNON successfully identified significantly clustered and differentially expressed Gene Ontology terms in breast cancer subtypes.
• The method detected functional annotations missed by standard FEA, complementing existing approaches.
• Results highlighted dysregulated and clustered functional annotations between HER2+, triple-negative, and hormone receptor-positive breast cancer subtypes.

Conclusions:

• PIGNON offers an alternative to traditional functional enrichment analyses.
• It provides a more comprehensive characterization of quantitative proteomics datasets.
• The method enhances the understanding of dysregulated functions and processes in biological samples.