SigHunt

SigHunt identifies genomic islands in eukaryotic genomes by detecting regions with atypical tetranucleotide (4-mer) composition indicative of horizontal gene transfer.

Key Features:

• Genomic Signature Analysis: calculates tetranucleotide (4-mer) frequencies using a sliding window across the genome to derive genomic signatures.
• Local Kernel Density Estimation: extends the neighborhood of each sliding window to compute local kernel density estimates for 4-mer frequencies.
• Discrete Interval Accumulative Score (DIAS): scores the number of 4-mers per window that deviate from local density credibility intervals to detect non-ameliorated integration events.
• Informative 4-mer Selection: selects informative 4-mers using a tetranucleotide quality score to improve sensitivity and specificity across organisms.
• Computational Efficiency: designed to scan large genomes with varying DNA compositions efficiently.

Scientific Applications:

• Genomic island detection in eukaryotes: identification of horizontally acquired DNA fragments that may carry novel genes or functions within eukaryotic genomes.
• Evolutionary genomics: analysis of horizontal gene transfer contributions to adaptability and evolution in eukaryotic lineages.

Methodology:

Compute tetranucleotide (4-mer) frequencies in sliding windows, estimate local kernel density of 4-mer frequencies across extended neighborhoods, score windows with Discrete Interval Accumulative Score (DIAS) against local density credibility intervals, and select informative 4-mers using a tetranucleotide quality score.