Targeting myocardial infarction-specific protein interactions using computational analyses
Background. Each year, over one million Americans experience myocardial infarction. Myocardial infarction is the cause of 70% of heart failure cases and the 5-year mortality rate for chronic heart failure is 50%. This high mortality is caused by the lacking of diagnostic and prognostic biomarkers. The purpose of this study was to develop a framework to better understand MI-specific PPI interaction network and identify MI-specific biomarkers.
Methods. We constructed an MI-specific PPI network by selecting 77 MI-related seed protein and link these seed proteins through their interactive proteins based on public protein interaction databases. An algorithm was developed to evaluate the significance of each protein-protein interaction based on their connective properties in the PPI network and their research intensity in PubMed abstract. We further employed 1000 protein networks, which was generated by randomly chosen 77 human proteins.
Results. Our MI-specific protein interaction network demonstrated statistical significance in their structural property compared to the random networks. The established MI-specific PPI network has less sub-networks and more links suggesting condensed protein interactions associated with disease. Higher measures of closeness centrality, clustering coefficient and degree centrality demonstrated a strong connectivity of hub proteins, which give us further confirmation about key proteins or biomarkers based on structural evaluation.