Whatizit

Whatizit identifies and annotates molecular biology terms in biomedical text and links them to bioinformatics database entries to support extraction and integration of biological information from scientific literature.

Key Features:

• Text Analysis Modules: A suite of modular text-mining modules extracts and analyzes information from scientific publications and Medline abstracts, identifying molecular biology terms.
• Database Linking: Links identified terms to UniProtKB/Swiss-Prot entries and Gene Ontology concepts to provide direct access to curated biological annotations.
• Annotation Identification: Recognizes annotation types including annotations produced by the EBIMed analysis pipeline for proteins.
• Server-based Scalability: Implements a server-based architecture to enable scalable processing of large literature datasets, including PubMed/Medline collections.

Scientific Applications:

• Literature-scale entity extraction: Extraction and normalization of molecular biology entities from large corpora of biomedical literature for downstream curation and analysis.
• Integration with curated databases: Facilitates linking of textual mentions to UniProtKB/Swiss-Prot and Gene Ontology entries to integrate unstructured text with structured biological annotations.

Methodology:

Server-based, modular architecture that applies text-mining modules to identify molecular biology terms in scientific publications and Medline abstracts, link them to UniProtKB/Swiss-Prot and Gene Ontology entries, and recognize EBIMed-derived protein annotations.