IBDDB

IBDDB consolidates curated and text-mined genetic and literature-derived biomedical entity data related to inflammatory bowel disease (including Crohn's disease and ulcerative colitis) to support analysis of experimentally validated gene–disease associations.

Key Features:

• Curated gene set: Contains 289 genes manually curated based on experimental validation.
• Text-mined PubMed knowledgebase: Integrates an extensive text-mined knowledgebase derived from PubMed records that are cleaned and curated by experts.
• Indexed records and dictionaries: Performs entity analyses across over one million indexed records using 11 curated dictionaries.
• Subject-specific concepts: Organizes information across 34 subject-specific concepts presented as columns for selective data display.
• Co-occurrence analysis: Enables investigation of biomedical entities and their co-occurrence with other entities extracted from the literature.
• Network visualization: Provides visualization of information in network formats to represent relationships and interactions.
• Data export: Supports selective data exportation in various file formats.
• Resource integration: Integrates multiple resources into a cohesive knowledgebase linking genetic and literature-derived data.

Scientific Applications:

• Gene–disease association studies: Supports analysis of experimentally validated genes associated with IBD, Crohn's disease, and ulcerative colitis.
• Hypothesis generation and cross-examination: Facilitates generation and cross-examination of novel hypotheses derived from literature co-occurrence patterns.
• Network analysis of interactions: Enables visualization-driven analysis of complex genetic and entity interaction networks.
• Literature-based relationship discovery: Supports discovery of biomedical entity relationships through large-scale text mining of PubMed.

Methodology:

Text-mining of PubMed records, expert-driven cleaning and manual curation yielding 289 experimentally validated genes, indexing of over one million records across 11 curated dictionaries, extraction of entity co-occurrence, organization into 34 subject-specific concept columns, and generation of network visualizations.