iHS

iHS detects signals of recent positive selection in human populations by quantifying extended haplotype homozygosity around single nucleotide polymorphisms and comparing ancestral and derived alleles.

Key Features:

• Extended haplotype homozygosity measurement: Quantifies the extent of extended haplotype homozygosity around individual SNPs.
• Ancestral versus derived allele comparison: Focuses on differences in haplotype structure between ancestral and derived alleles at each SNP.
• iHS score standardization: Standardizes iHS scores empirically against the distribution of observed scores across SNPs with similar derived allele frequencies.
• Genome-wide SNP scanning: Applies iHS across genome-wide SNP datasets to detect candidate loci under selection.
• Population-scale analyses: Supports application to diverse population datasets, as demonstrated using the International HapMap Project.
• Tag SNP selection: Identifies sets of SNPs that tag the strongest signals of recent selection in each population.
• Detection of selective sweeps: Highlights loci where high-frequency derived alleles exhibit extended homozygosity consistent with selective sweeps.

Scientific Applications:

• Detect recent positive selection: Identifies genomic regions that have undergone recent positive selection in human populations.
• Study human adaptation: Provides insights into human adaptation to local environmental conditions by pinpointing candidate adaptive loci.
• Guide mapping of complex traits: Prioritizes loci for mapping studies aimed at understanding phenotypic variation and complex trait architecture.
• Compare population-specific signals: Reveals region-specific and shared signals of selection across continental groups, including signals detected in the Yoruba and other populations often overlooked by lower-resolution studies.

Methodology:

Compute extended haplotype homozygosity around SNPs, compare ancestral and derived alleles to derive an iHS score, standardize iHS empirically across SNPs binned by derived allele frequency, and apply the statistic genome-wide to SNP datasets such as those from the International HapMap Project.