pGenN

pGenN performs gene normalization for plant species by detecting gene and protein mentions in scientific literature and mapping them to standardized database identifiers to structure plant gene information.

Key Features:

• Gene Mention Detection: Uses dictionary-based approaches to identify gene and protein mentions in plant-related texts.
• Species Assignment and Intra-species Normalization: Assigns detected mentions to specific species and maps nomenclatures and synonyms to standardized identifiers.
• Advanced Heuristics: Incorporates newly developed heuristics to improve detection, species assignment, and normalization accuracy.
• Performance Metrics: Evaluated on an expertly annotated corpus of 104 plant-relevant abstracts with an F-value of 88.9% (Precision: 90.9%, Recall: 87.2%), outperforming state-of-the-art systems from the BioCreative III challenge.
• Large-scale Processing: Processed over 440,000 plant-related Medline abstracts.

Scientific Applications:

• Database Curation: Improves gene annotation coverage by linking literature mentions to standardized identifiers.
• Text Mining Systems: Enhances extraction of gene/protein information from unstructured plant literature.
• Research and Development: Provides standardized gene and protein mappings to support literature review and downstream plant genomics analyses.

Methodology:

Uses dictionary-based gene/protein mention detection, species assignment, intra-species normalization mapping synonyms to standardized identifiers, and newly developed heuristics for detection, assignment, and normalization; evaluated on an annotated corpus of 104 plant abstracts and applied to >440,000 plant-related Medline abstracts.