LitMiner

LitMiner extracts literature-derived relationships among genes, chemical compounds, diseases, and tissues from PubMed abstracts using statistical co-occurrence analysis to support gene regulatory network inference and promoter modeling.

Key Features:

• Co-occurrence Analysis: Performs statistical co-occurrence (co-citation) analysis on PubMed abstracts to identify associations among biomedical terms.
• Term Annotation: Annotates key terms within PubMed abstracts, including genes, chemical compounds, diseases, and tissues, for downstream analysis.
• Categorization: Classifies identified relationships into the four domains: genes, chemical compounds, diseases, and tissues.
• Predictive Regulatory Network Inference: Infers potential regulatory networks and gene–gene or gene–chemical interactions from co-occurrence patterns.
• Primary Gene List Generation: Generates primary gene lists from co-occurrence results for promoter modeling and experimental follow-up.
• Integration with WikiGene: Enables refinement and verification of predicted associations via the WikiGene wiki-based curation platform.

Scientific Applications:

• Disease-related regulatory network reconstruction: Reconstructs disease-related regulatory networks using promoter modeling informed by literature-derived primary gene lists.
• Hypothesis generation for gene regulation: Produces hypotheses about gene regulation mechanisms based on annotated co-occurrence associations.
• Identification of candidate associations: Identifies candidate links among genes, chemical compounds, diseases, and tissues for targeted experimental validation.

Methodology:

Annotation of key terms within PubMed abstracts followed by statistical co-occurrence/co-citation analysis, with predicted associations subject to refinement via WikiGene curation.