Improving microbial genome annotations in an integrated database context

• Published in: PloS one Vol. 8; no. 2; p. e54859
• Main Authors: Chen, I-Min A, Markowitz, Victor M, Chu, Ken, Anderson, Iain, Mavromatis, Konstantinos, Kyrpides, Nikos C, Ivanova, Natalia N
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 12.02.2013; Public Library of Science (PLoS)
• Subjects: Acids; Analysis; Annotations; Automation; Bioinformatics; Biology; Comparative analysis; Completeness; Computer Science; Consistency; Databases, Genetic; Genes; Genomes; Genomics; Laboratories; Metabolism; Metadata; Microbiology; Microorganisms; Molecular Sequence Annotation - methods; Phenotype; Physiology; Proteins; Respiration; Semantics; California
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0054859

Effective comparative analysis of microbial genomes requires a consistent and complete view of biological data. Consistency regards the biological coherence of annotations, while completeness regards the extent and coverage of functional characterization for genomes. We have developed tools that allow scientists to assess and improve the consistency and completeness of microbial genome annotations in the context of the Integrated Microbial Genomes (IMG) family of systems. All publicly available microbial genomes are characterized in IMG using different functional annotation and pathway resources, thus providing a comprehensive framework for identifying and resolving annotation discrepancies. A rule based system for predicting phenotypes in IMG provides a powerful mechanism for validating functional annotations, whereby the phenotypic traits of an organism are inferred based on the presence of certain metabolic reactions and pathways and compared to experimentally observed phenotypes. The IMG family of systems are available at http://img.jgi.doe.gov/.