Multi-source and ontology-based retrieval engine for maize mutant phenotypes.

Area of Science:

• Genomics
• Bioinformatics
• Plant Science

Background:

• Model organism databases, like plant genome databases, house vast heterogeneous data including sequences, gene products, and phenotypic information.
• Existing search capabilities in these databases are often limited, lacking advanced text description querying and ranking.
• Plant research communities require novel methods to simultaneously search linked text descriptions across databases.

Purpose of the Study:

• To develop an advanced search engine for mutant phenotypes within MaizeGDB.
• To address limitations in current plant genomic database search functionalities.
• To leverage ontologies for enhanced retrieval of phenotypic data.

Main Methods:

• Developed a unique search engine integrating diverse text description sources from MaizeGDB (mutant phenotypes, loci, gene products).
• Exploited the content and structure of domain ontologies, specifically the Plant and Gene Ontologies.
• Designed a retrieval engine to expand and enrich search results in major plant genomic databases.

Main Results:

• Successfully integrated multiple text description sources for comprehensive mutant phenotype searches.
• The engine is the first to utilize ontologies to enhance retrieval in plant genomic databases.
• Demonstrated an innovative approach to querying and ranking textual data for phenotypic information.

Conclusions:

• The developed search engine significantly improves mutant phenotype information retrieval in MaizeGDB.
• The integration of ontologies offers a novel and effective method for enriching search results in plant genomics.
• This approach provides a valuable tool for the plant research community, enhancing data accessibility and discovery.