Diffusion Alignment Coefficient (DAC): A Novel Similarity Metric for Protein-Protein Interaction Network

Area of Science:

• Bioinformatics
• Computational Biology
• Systems Biology

Background:

• Protein-protein interaction (PPI) networks are crucial for predicting protein functions.
• Existing graph theory and diffusion-based methods assume topological properties correlate with function.
• These methods often suffer from topological information loss.

Purpose of the Study:

• To improve protein function prediction by developing a novel diffusion-based alignment technique.
• To overcome the limitations of existing methods by reducing topological information loss.
• To introduce a new measure for node function similarity in PPI networks.

Main Methods:

• Development of the Diffusion Alignment Coefficient (DAC) algorithm.
• Integration of diffusion, longest common subsequence, and longest common substring techniques.
• Application of DAC to PPI networks in S.cerevisiae and Homo Sapiens.

Main Results:

• DAC demonstrated superior performance compared to existing methods.
• Improved functional categorization accuracy for MIPS and MSigDB Collections hallmark gene sets.
• Validation of the positive impact of spatial information on detecting functionally similar nodes.

Conclusions:

• The DAC algorithm effectively measures node function similarity in PPI networks.
• Incorporating spatial information enhances the prediction of functionally related proteins.
• This study presents a novel approach to protein function prediction using network alignment.