BioCAT

BioCAT identifies potential producers of non-ribosomal peptides (NRPs) by aligning nonribosomal peptide synthetase (NRPS) module specificities and peptide chemical structures to genomic and metagenomic datasets.

Key Features:

• antiSMASH integration: Incorporates features from antiSMASH for gene cluster detection and annotation.
• rBAN retrosynthesis integration: Integrates rBAN retrosynthesis for peptide chemical structure analysis.
• Position-specific score matrix (PSSM): Implements a PSSM that captures NRPS module specificities to improve alignment sensitivity.
• Ensemble scoring: Uses an ensemble model to calculate final alignment scores for producer prediction.
• Gene cluster and peptide structure improvements: Includes enhancements to gene cluster and peptide chemical structure analyses compared to previous approaches.
• Large-dataset automation: Automates identification of NRP producers from large datasets such as metagenomes.
• Benchmarking: Validated against a manually curated database of NRP producers and benchmarked with GARLIC, demonstrating superior performance in several external examples.

Scientific Applications:

• NRP producer identification: Screening genomes and metagenomes to identify natural producers of non-ribosomal peptides.
• NRPS gene cluster analysis: Annotating and comparing nonribosomal peptide synthetase gene clusters based on module specificities.
• Peptide structure inference: Linking retrosynthesized peptide chemical structures to candidate producer organisms.
• Method benchmarking: Comparative evaluation of producer-identification methods, including benchmarking against GARLIC.

Methodology:

Integrates antiSMASH features and rBAN retrosynthesis, implements a position-specific score matrix (PSSM) for NRPS module specificities, employs an ensemble model to calculate final alignment scores, and was tested against a manually curated NRP producer database and benchmarked with GARLIC.