Related Experiment Videos
Gene annotation from scientific literature using mappings between keyword systems.
Antonio J Pérez1, Carolina Perez-Iratxeta, Peer Bork
1University of Málaga, Facultad de Ciencias, Departmento de Genetica, Group of Bioinformatics, Campus Universitario de Teatinos, 29071 Málaga, Spain.
Bioinformatics (Oxford, England)
|April 3, 2004
Summary
This study introduces an automated method to extract keywords from scientific abstracts for gene annotation. The system improves gene data accessibility and computational tool efficiency for researchers.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- Gene databases benefit from keyword descriptions for non-specialists and computational tools.
- Associating keywords with genes/proteins is challenging, requiring literature review.
Purpose of the Study:
- To develop an automated procedure for deriving keywords from scientific abstracts linked to genes.
- To enhance the efficiency of gene data searching and understanding.
Main Methods:
- Automated extraction of term mappings between databases using fuzzy associations.
- Application to SWISS-PROT and MEDLINE databases for gene annotation.
- Testing keyword derivation from abstracts for gene entries.
Main Results:
- The system achieved better performance annotating SWISS-PROT keywords (47% recall, 68% precision) compared to Gene Ontology terms (8% recall, 67% precision).
- Demonstrated a procedure to derive gene-related keywords from scientific literature.
Conclusions:
- The developed system effectively annotates genes with keywords extracted from abstracts.
- The fuzzy association model offers a generalizable approach for linking diverse databases.
- Publicly accessible web server facilitates sequence annotation.