Novel Bioinformatics Approach Reveals Pathogenic Mechanisms In Cerebral Ischemia - A Step Towards Preclinical Stroke Information Management System

Introduction: A plethora of pathophysiological mechanisms comes into play after ischemic stroke. They contribute to both primary and secondary insult and determine the functional outcome of the patient. Several drug treatable molecular pathways for stroke pathogenesis have been described over decades of research in cerebrovascular ischemic injury. However, none of these pathways acts independently and serves as a sole determinant of long-term outcome. This necessitates a standardized approach that merges previous efforts and reveals the full spectrum of interacting pathways in stroke pathogenesis. Objective: The aim of this study is to use advanced annotation and text mining strategies combined with systems biology resources to compile a spectrum of molecular pathways in the network of stroke pathogenesis and highlight therapeutic targets. Methods: We constructed and utilized a novel annotation tool for a semi-automatic annotation of literature for proteins/genes/pathways reported to be involved in stroke pathogenesis and recovery. Frequency of both occurrence and co-occurrence of extracted entities was computed. Functional annotation and clustering was performed to form a brain ischemic meta-proteome. Network of annotated protein-protein interaction was visualized and analyzed for enriched pathways and prominent hub proteins that are potential targets. Results: A total of 80,941 PubMed articles on brain ischemia were annotated and an additional 1,240 articles were obtained from alternative sources. Around 1,000 proteins were reported in stroke pathogenesis with a frequency of occurrence ranging between 2 and 879. The extracted brain ischemic proteome was found to match genetic association studies. Pathway analysis revealed that coagulation and complement pathway and calcium signaling were two most prominent pathways. Inflammatory mechanisms were the most significantly enriched processes in the brain ischemia proteome. Several hubs were identified in the network including TNF-α, VEGFA, NOS3, IL1B, PTGS2, ICAM1, IL6, and NGF. Conclusion: We present a meta-proteome for brain ischemia that reveals different interacting pathways in stroke pathogenesis and serves as a step towards a Preclinical Stroke Information Management System.