MetaGeneAnnotator

MetaGeneAnnotator predicts prokaryotic genes from environmental genome shotgun and metagenomic sequences to enable accurate gene annotation and translation start site identification across bacterial and archaeal genomes.

Key Features:

• Comprehensive gene prediction: Predicts genes from single or multiple anonymous genomic sequences of varying lengths, including bacterial, archaeal, horizontally transferred, and prophage genes.
• Statistical models and self-training: Integrates statistical models for standard bacterial/archaeal genes and prophage genes together with a self-training model that adapts to input sequences.
• Ribosomal binding site (RBS) analysis: Analyzes species-specific RBS patterns to improve prediction of translation start sites.
• Performance on short sequences: RBS model achieves reported 96% sensitivity and 93% specificity for predicting translation starts, effective on sequences as short as 700 base pairs.

Scientific Applications:

• Genome annotation: Enables accurate annotation of prokaryotic genomes including draft and fragmented sequences.
• Metagenomic analysis: Facilitates gene prediction from environmental shotgun and metagenomic datasets.
• Horizontal gene transfer and prophage detection: Improves identification of atypical genes such as horizontally transferred and prophage genes.
• Microbial diversity and function studies: Supports exploration of genetic diversity and functional potential within prokaryotic communities.

Methodology:

Integration of statistical models for bacterial, archaeal and prophage genes, a self-training algorithm that adapts to input sequences, and analysis of ribosomal binding sites for translation start prediction.