Platon

Platon identifies plasmid-derived contigs in bacterial draft assemblies by distinguishing plasmid-borne from chromosome-borne sequences using replicon distribution patterns to enable large-scale plasmid characterization and surveillance of antimicrobial resistance.

Key Features:

• Taxon-independent plasmid identification: Performs plasmid sequence identification with high sensitivity and specificity across diverse bacterial taxa.
• Replicon Distribution Score (RDS): Uses a metric based on replicon distributions of protein-coding genes to discriminate plasmid-borne and chromosome-borne contigs, achieving an accuracy of 96.6%.
• Enhanced performance through heuristics: Integrates heuristics that exploit plasmid-specific contig characteristics with RDS to produce a high-throughput workflow with 97.5% accuracy and an F1 score of 82.6%.
• Comparative benchmarking: Demonstrated superior or equivalent performance to PlasFlow, PlasmidFinder, and PlaScope on sequenced Escherichia coli isolates.

Scientific Applications:

• Plasmid extraction from draft assemblies: Facilitates extraction and analysis of plasmid sequences from draft genome assemblies.
• Study of bacterial adaptation: Supports investigations of plasmid-mediated mechanisms of bacterial adaptation, including dissemination of antimicrobial resistance genes and virulence factors.
• Large-scale plasmid epidemiology: Enables large-scale genomic studies of plasmid origin, distribution, and genetic repertoire to inform surveillance and research on antibiotic resistance.

Methodology:

Computes a Replicon Distribution Score from distributions of protein-coding genes across replicons, applies statistical discrimination thresholds and plasmid-specific heuristics within a high-throughput workflow to classify contigs as plasmid-borne or chromosome-borne.