GIDL: a rule based expert system for GenBank Intelligent Data Loading into the Molecular Biodiversity Database

Area of Science:

• Bioinformatics
• Biodiversity Informatics
• Computational Biology

Background:

• Growing need to integrate molecular and traditional biodiversity data.
• Information technology can bridge molecular data from public databases and biodiversity studies.
• Current approaches lack robust integration of diverse biodiversity datasets.

Purpose of the Study:

• Develop a bioinformatic infrastructure for integrating public and private biodiversity data.
• Create an Intelligent Data Loader (GIDL) coupled with a Molecular Biodiversity Database.
• Organize and locally store sequence and annotation data from GenBank using an ontological framework.

Main Methods:

• Designed a relational database (Molecular Biodiversity Database) with four areas: MolecularData, Experiment, Collection, and Taxonomy.
• Utilized the Chado relational schema and Sequence Ontology for the MolecularData area.
• Developed an Intelligent Data Loader (IDL) using Extract, Transform, Load (ETL) processes, including a Parser, Reasoner, and DBFiller, leveraging Semantic Web technologies.

Main Results:

• Successfully loaded over 3 million GenBank entries (Virus, Plant, Invertebrate) into the Molecular Biodiversity Database.
• GIDL effectively parses GenBank data, infers biological knowledge using the Sequence Ontology, and populates the database.
• The system demonstrates efficient data integration and organization.

Conclusions:

• The GIDL system, integrating Sequence Ontology and Chado schema, offers superior query capabilities compared to standard retrieval systems.
• Provides a powerful bioinformatic infrastructure for biodiversity data management and analysis.
• Facilitates a more comprehensive understanding of biodiversity through integrated data.