IslandCompare

IslandCompare predicts and compares genomic islands (GIs) across multiple bacterial genomes to identify horizontally acquired regions that encode virulence factors and antimicrobial resistance (AMR) determinants.

Key Features:

• Composition-based GI prediction: Uses Sigi-HMM and IslandPath-DIMOB, both based on nucleotide composition biases, to predict genomic islands.
• Blast-based consistency step: Applies a blast-based consistency step to enhance cross-genome prediction reliability.
• Core-genome phylogeny mapping: Constructs a bacterial core-genome phylogeny and arranges genomes at tree leaves for comparative mapping of features.
• AMR and GI overlays: Overlays GI predictions and AMR determinants sourced from the Comprehensive Antibiotic Resistance Database (CARD) onto genome representations.
• Genomic similarity and clustering: Computes regions of genomic similarity and groups GIs across genomes into sequence-similarity clusters to facilitate comparative analysis.

Scientific Applications:

• Bacterial evolution: Comparative GI analysis to study horizontal gene transfer and the dynamics of mobile genetic elements across lineages.
• Virulence research: Identification and comparison of GI-encoded virulence factors to investigate pathogenicity determinants.
• Antimicrobial resistance studies: Mapping of AMR genes from CARD within GIs to trace emergence and spread of resistance determinants.
• Outbreak and sub-clade analysis: Tracking acquisition and loss of mobile GIs within sub-clades to inform outbreak dynamics and lineage-specific adaptation.

Methodology:

Runs Sigi-HMM and IslandPath-DIMOB for composition-based GI prediction, applies a blast-based consistency step for cross-genome reliability, constructs a core-genome phylogeny with genomes arranged at tree leaves, overlays GI predictions and CARD AMR determinants, computes regions of genomic similarity, and clusters GIs across genomes by sequence similarity.