Candidate gene prioritization for chronic obstructive pulmonary disease using expression information in protein–protein interaction networks

Abstract

Background: Identifying or prioritizing genes for chronic obstructive pulmonary disease (COPD), one type of complex disease, is particularly important for its prevention and treatment. Methods: In this paper, a novel method was proposed to Prioritize genes using Expression information in Protein–protein interaction networks with disease risks transferred between genes (abbreviated as PEP). A weighted COPD PPI network was constructed using expression information and then COPD candidate genes were prioritized based on their corresponding disease risk scores in descending order. Results: Further analysis demonstrated that the PEP method was robust in prioritizing disease candidate genes, and superior to other existing prioritization methods exploiting either topological or functional information. Top-ranked COPD candidate genes and their significantly enriched functions were verified to be related to COPD. The top 200 candidate genes might be potential disease genes in the diagnosis and treatment of COPD. Conclusions: The proposed method could provide new insights to the research of prioritizing candidate genes of COPD or other complex diseases with expression information from sequencing or microarray data.