multiple tools explore

multiple tools explore (MeTeOR) leverages text mining of PubMed and MeSH co-occurrence networks to extract literature-derived gene, disease, and drug pathways for disease- and drug-specific pathway discovery.

Key Features:

• Network Representation: MeTeOR constructs a co-occurrence network from PubMed using Medical Subject Headings (MeSH) terms that link genes, diseases, and drugs.
• Community Detection: The method applies community detection algorithms to identify 3444 functional gene groups representing specific biological pathways associated with diseases and drugs.
• Pathway Synthesis: MeTeOR synthesizes extensive pathways by aggregating community structures to recapitulate known pathways and predict novel pathways using time-stamped experimental data.
• Predictive Power: The approach recovers mutual drug targets (AUROC=0.75) and shared pathogenic genes across diseases (AUROC=0.82).
• Comorbidity and Side Effects Analysis: Diseases that share communities exhibit high comorbidity and drugs within the same community often have common side effects, indicating related mechanisms.

Scientific Applications:

• Disease Gene Discovery: Identifying shared pathogenic genes across diseases to uncover genetic contributors to disease etiology.
• Drug Repurposing: Prioritizing existing drugs for new therapeutic uses by predicting mutual drug targets and shared pathways.
• Pathway Analysis: Exploring comprehensive literature-derived biological pathways to elucidate disease mechanisms and treatment effects.

Methodology:

MeTeOR mines PubMed via text mining to construct a MeSH-term co-occurrence network and applies community detection algorithms to synthesize functional gene groups representing biological pathways; it uses time-stamped experimental data to support pathway prediction.