A New Framework for Pinpointing Crucial Proteins in Protein-Protein Interaction Networks

Abstract

Identifying crucial proteins in protein-protein interaction (PPI) networks is essential for understanding biological systems. However, the ambiguity of interaction strength hinders accurate identification of key proteins. To address this issue, this study proposes a new framework that introduces Bio-Link Strength, a fuzzy membership function that utilizes fuzzy set theory to quantifies interaction strength with continuous values between 0 and 1. Framework extends four commonly used traditional measures (degree, closeness, betweenness, and eigenvector) to fuzzy measures (fuzzy connectivity, fuzzy approachability, fuzzy bridge, and fuzzy influence centrality), enabling effective identification of crucial proteins. The efficacy of the proposed framework has been assessed on different commonly used real-world PPI network datasets (Saccharomyces cerevisiae, Escherichia coli, and Drosophila melanogaster), to prove the framework’s scalability. Results show that proposed membership function effectively assesses protein interaction strength, with a strong positive Spearman’s correlation between fuzzy and traditional measures. Furthermore, Gene Ontology analysis confirms the importance of top proteins identified by our fuzzy measures. Notably, our fuzzy connectivity and influence centrality measures outperform their traditional counterparts and other proposed fuzzy measures in identifying crucial proteins.

Keywords

• Crucial proteins
• protein-protein interaction networks
• edge strength
• fuzzy membership function
• fuzzy centrality measures
• fuzzy biological networks