GIST

GIST predicts Genomic Islands (GIs) by integrating multiple GI-detection algorithms to identify genomic regions acquired via horizontal gene transfer (HGT) that may encode pathogenicity, antibiotic resistance, or metabolic capabilities.

Key Features:

• Integrated toolset: Incorporates AlienHunter, IslandPath, Colombo SIGI-HMM, INDeGenIUS, and Pai-Ida, and supports using these tools independently or sequentially for GI detection.
• Ensemble optimization (EGID): Uses the EGID program to integrate outputs from the individual tools to refine and consolidate GI predictions.
• Automated data retrieval: Supports automatic download of input genomes from the FTP server of the National Center for Biotechnology Information (NCBI).
• Implementation: Implemented in Java and compiled/executed on Linux/Unix operating systems.

Scientific Applications:

• Microbial genomics: Identification of horizontally acquired genomic regions that contribute to strain-specific functions and adaptation.
• Evolutionary biology: Analysis of HGT events and their role in microbial evolution and genome plasticity.
• Infectious disease research: Detection of genomic regions associated with virulence and antibiotic resistance for studies of pathogenesis and resistance spread.

Methodology:

GIST applies AlienHunter (codon usage anomaly detection), IslandPath (compositional and gene cluster analyses), Colombo SIGI-HMM (hidden Markov models for atypical dinucleotide frequencies), INDeGenIUS (integration of multiple indicators), and Pai-Ida (pathogenicity island detection), and integrates their outputs using the EGID ensemble optimizer.