SGI

SGI detects positive and negative selection at single codon sites in protein-coding sequences by comparing synonymous and nonsynonymous substitutions across phylogenetic branches reconstructed from synonymous substitution counts.

Key Features:

• Per-codon neutrality testing: Tests neutrality at each codon by comparing numbers of synonymous (silent) and nonsynonymous (amino-acid altering) substitutions across the phylogeny.
• Multiple sequence alignment and tree reconstruction: Utilizes multiple sequence alignments and phylogenetic trees reconstructed based on synonymous substitution counts to place substitutions on branches.
• Simulation validation: Validated by computer simulations that demonstrate accurate estimation of substitution rates per site when the number of substitutions per branch remains relatively low.
• False-positive characteristics: Exhibits a generally low false-positive rate for detecting selective forces.
• Parameter-dependent true-positive rate: True-positive detection rate varies with parameter values and increases with greater selective strength and larger total branch length measured in synonymous substitutions per site.
• Performance threshold: Effectively identifies positively selected codon sites when the ratio of nonsynonymous to synonymous substitution rates is 5.0 and total branch length (in synonymous substitutions per site) is ≥ 2.5.

Scientific Applications:

• HLA gene ARS analysis: Applied to human leukocyte antigen (HLA) genes, detecting positive selection predominantly at antigen recognition sites (ARSs) and identifying additional positively selected non-ARS sites supported by structural analyses.
• HIV envelope protein: Identified positively selected codon sites in the envelope protein of human immunodeficiency virus.
• Influenza hemagglutinin: Identified positively selected codon sites in influenza virus hemagglutinin protein.
• Functional prediction from selection patterns: Used to predict functional aspects of amino acid sites as sequence data accumulate.

Methodology:

Uses multiple sequence alignments and phylogenetic trees reconstructed from synonymous substitution counts; compares numbers of synonymous and nonsynonymous substitutions across phylogenetic branches to test neutrality at each codon site; validated by computer simulations showing accurate substitution-rate estimates when substitutions per branch are relatively low and reports total branch length in synonymous substitutions per site.