An integrated approach to identify protein complex based on best neighbour and modularity increment

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Protein-Protein Interaction (PPI) networks are crucial for understanding cellular functions.
• Existing modularity measures for PPI networks have limitations in capturing both global and local network structures.
• Accurate identification of protein complexes is essential for deciphering biological pathways.

Purpose of the Study:

• To introduce a novel hybrid modularity measure, Local-Global Quantification (LGQ), addressing limitations of existing methods.
• To develop an advanced protein complex mining algorithm, Best Neighbour and Local-Global Quantification (BN-LGQ).
• To enhance the accuracy and biological relevance of protein complex prediction in PPI networks.

Main Methods:

• Proposed the Local-Global Quantification (LGQ) measure, integrating global and local modularity with an adjustable parameter.
• Developed the Best Neighbour and Local-Global Quantification (BN-LGQ) algorithm, combining best neighbor node search and modularity increment.
• Utilized modularity increment as a metric for expanding protein complexes within the PPI network.

Main Results:

• BN-LGQ demonstrated superior accuracy in predicting protein complexes compared to MCL and MCODE algorithms.
• The algorithm achieved a higher match rate with known reference protein complexes.
• BN-LGQ effectively identified protein complexes with enhanced biological significance.

Conclusions:

• The LGQ measure provides a balanced approach to detecting protein complexes by considering both global and local network properties.
• The BN-LGQ algorithm offers an effective and accurate method for mining protein complexes from PPI networks.
• BN-LGQ advances the field by improving the biological interpretability and discovery of protein complexes.