emiRIT

emiRIT provides extracted microRNA (miRNA) information from Medline abstracts using text-mining to support analysis of miRNA roles in gene regulation, biological processes, pathways, extracellular locations, and disease (notably cancer).

Key Features:

• Extensive Data Collection: Compiled 149,233 miRNA–PubMed ID pairs from Medline records spanning January 1997 to May 2020.
• Categorized Relations: Annotated relations include miRNA–gene regulation (60,491 relations), miRNA–disease associations (12,300 relations for cancer), miRNA–biological processes and pathways (23,390 relations), and circulatory miRNAs in extracellular locations (3,782 relations).
• Advanced Text-Mining Tools: Employs both publicly available and custom-developed text-mining tools to extract biological entities and their relationships with miRNAs from Medline abstracts.
• Normalization of Extracted Entities: Performs normalization of extracted entities to ensure consistency and facilitate integration into the resource.
• Database Integration: Integrates extracted and normalized miRNA information into a structured database infrastructure for downstream analysis.

Scientific Applications:

• Centralized miRNA knowledgebase: Serves as a consolidated repository of literature-derived miRNA associations and evidence.
• Gene regulation analysis: Enables exploration of miRNA–gene regulatory relationships derived from the literature.
• Disease investigation (cancer): Provides literature-based miRNA–disease association data with a specific corpus for cancer-related relations.
• Pathway and process studies: Supports analysis of miRNA involvement in biological processes and pathways documented in Medline.
• Extracellular miRNA research: Supplies curated evidence on circulatory/extracellular miRNAs relevant to biomarker studies.

Methodology:

Systematic extraction of miRNA information from Medline abstracts using advanced text-mining techniques and both public and custom tools; normalization of extracted entities to ensure consistency and facilitate integration; compilation of 149,233 miRNA–PubMed ID pairs (January 1997–May 2020) and categorization of relations into miRNA–gene, miRNA–disease (cancer), miRNA–process/pathway, and extracellular miRNA relations; integration of the processed data into a database infrastructure.