decRiPPter

decRiPPter identifies candidate ribosomally synthesized and post-translationally modified peptide (RiPP) biosynthetic gene clusters (BGCs) and novel RiPP classes by combining a Support Vector Machine (SVM) with pan-genomic analyses to detect precursor peptides and operon-like accessory-genome loci beyond traditional sequence similarity searches.

Key Features:

• Machine Learning Integration: Employs a Support Vector Machine (SVM) to detect candidate RiPP precursor peptides within genome sequences.
• Pan-genomic Analysis: Performs pan-genomic analyses to determine which candidate RiPPs are encoded in operon-like structures within the accessory genome of a genus.
• Operon Identification and Prioritization: Identifies operons with potential BGCs and prioritizes them based on novel enzymology and conservation patterns of gene clusters and precursor peptides across species.
• Discovery beyond Sequence Similarity: Integrates multiple pathway discovery criteria to uncover RiPP families that evade traditional sequence similarity searches.
• Empirical Application to Streptomyces: Applied across 1,295 Streptomyces genomes to identify 42 candidate RiPP families.
• Novel Enzymology and Product Classes: Enabled the identification of a novel lanthipeptide subfamily (Class V) including two previously unidentified modifying enzymes responsible for forming lanthionine bridges.

Scientific Applications:

• RiPP Biosynthetic Landscape Expansion: Enables discovery and classification of previously unrecognized RiPP families and modifying enzymes.
• Biosynthetic Gene Cluster Prioritization: Provides criteria to prioritize candidate BGCs for biochemical characterization and functional studies based on enzyme novelty and conservation.
• Natural Product Discovery: Supports identification of novel natural products, including new lanthipeptide subclasses, for downstream chemical and biological investigation.

Methodology:

Utilizes a Support Vector Machine (SVM) to identify candidate RiPP precursors; conducts pan-genomic analyses to locate operon-like structures in the accessory genome; prioritizes genomic regions based on novel enzymatic functions and conservation patterns; integrates multiple pathway discovery criteria to go beyond sequence similarity searches and applied these methods across 1,295 Streptomyces genomes to identify 42 candidate RiPP families, including a Class V lanthipeptide subfamily with two enzymes forming lanthionine bridges.