EnzymeDetector

EnzymeDetector consolidates enzymatic function annotations across genomes by integrating multiple annotation databases and sequence-based prediction methods to improve annotation consistency.

Key Features:

• Comparative Analysis: Systematically compares enzyme annotations from BRENDA, UniProt, KEGG, PATRIC, and NCBI RefSeq and identifies inconsistencies across prokaryotic genomes, with studies reporting about 70% discrepancies among the main resources.
• Integrated Annotation Pipeline: Supplements existing database annotations with predictions from sequence similarity analysis, organism-specific enzyme information from BRENDA, and sequence pattern searches to provide a comprehensive view of enzyme annotations.
• Customizable Quality Criteria: Enables customization of weighting schemes and cut-offs for different prediction methods to adjust annotation scoring and confidence assessment.
• Error Detection and Resolution: Summarizes evidence across annotation sources and assigns low relevance scores to annotations supported by only a single database to highlight potential errors.

Scientific Applications:

• Metabolic Engineering: Provides accurate enzyme function data for design and optimization of metabolic pathways.
• Genomic Research: Resolves inconsistencies in enzyme predictions to improve genome annotation projects.
• Systems Biology: Supports reconstruction and analysis of metabolic models by supplying consolidated enzyme annotations.

Methodology:

Integration and comparison of annotations from BRENDA, UniProt, KEGG, PATRIC, and NCBI RefSeq; sequence similarity analysis including BLAST searches against Swiss-Prot and BrEPS; incorporation of manually curated organism-specific enzyme information from BRENDA; and sequence pattern searches.