SigHunt: horizontal gene transfer finder optimized for eukaryotic genomes

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genomic islands (GIs) are DNA fragments acquired via horizontal gene transfer, often conferring novel functions.
• Detecting GIs in prokaryotes is established, but eukaryotic genome complexity hinders direct application of existing methods.
• Current eukaryotic GI detection relies on labor-intensive phylogenetic analyses.

Purpose of the Study:

• To present SigHunt, a novel surrogate method for identifying putative genomic islands in eukaryotic genomes.
• To offer a computationally efficient alternative to traditional phylogenetic searches for GI detection.

Main Methods:

• Utilizes a sliding window approach to calculate tetranucleotide (4-mer) frequencies, creating a genomic signature.
• Employs a local kernel density estimate to assess 4-mer frequency deviations.
• Introduces a discrete interval accumulative score (DIAS) to quantify deviations and a tetranucleotide quality score for selecting informative 4-mers.

Main Results:

• SigHunt demonstrates computational efficiency in detecting GIs in eukaryotic genomes.
• The method successfully identifies GIs representing non-ameliorated integration.
• SigHunt is suitable for scanning genomes with varying DNA compositions.

Conclusions:

• SigHunt provides an effective and efficient computational tool for identifying genomic islands in eukaryotic genomes.
• The method's reliance on DNA composition analysis makes it adaptable to diverse eukaryotic organisms.
• This approach facilitates the study of horizontal gene transfer and its impact on eukaryotic genome evolution.