EpiGraphDB

EpiGraphDB integrates diverse biomedical and epidemiological relationships into a graph database to support systematic causal inference across phenotypes using Mendelian randomization.

Key Features:

• Graph Database Structure: Organizes and represents complex biomedical and epidemiological relationships in a graph-based schema for efficient querying and relationship traversal.
• Mendelian Randomization: Systematically applies Mendelian randomization across numerous phenotypes to generate causal estimates linking genetic variants to health outcomes.
• Data Integration and Curation: Amalgamates and curates data from multiple bioinformatic sources to provide comprehensive coverage of biomedical entities and relationships.
• Analytical Platform: Performs systematic automated analyses to support reproducible causal inference and large-scale epidemiological studies.
• Pleiotropy Evaluation: Evaluates potential pleiotropic relationships to reduce mis-inference in causal analyses.
• Drug Target Identification: Incorporates protein–protein interaction data to identify and prioritize candidate therapeutic targets.
• Triangulation of Evidence: Integrates causal inference results with relationships mined from biomedical literature to corroborate findings across data sources.

Scientific Applications:

• Pleiotropy Evaluation: Assesses pleiotropic effects to improve interpretation of genetic associations in causal analyses.
• Drug Target Identification: Identifies potential therapeutic targets by combining protein–protein interaction data with causal genetic evidence.
• Triangulation of Evidence: Corroborates findings by combining Mendelian randomization-derived causal estimates with literature-mined biomedical relationships.

Methodology:

Data from multiple bioinformatic sources are curated and integrated into a graph database; relationships (including protein–protein interactions and literature-derived links) are mined, and Mendelian randomization is systematically applied across phenotypes to generate causal estimates.