introgress

introgress estimates locus-specific genomic clines and analyzes gene introgression across hybrid zones to detect selection and estimate hybrid indices.

Key Features:

• Genomic clines estimation: Employs multinomial regression to estimate genomic clines across admixture gradients and quantify introgression at individual loci relative to the genomic background.
• Detection of selection: Identifies loci with introgression patterns that significantly deviate from neutral expectations, indicating potential selection pressures.
• Simulation-based validation: Uses simulations to assess power to detect moderate to strong selection across various selection forms and reports low false positive rates except under pronounced genetic drift (e.g., small population sizes or low migration rates).
• Detection range for selection: Can detect moderate selection against heterozygotes up to 50 centimorgans from a focal locus while showing no effect at unlinked loci.
• Versatile data compatibility: Supports co-dominant, dominant, and haploid marker data and does not require fixed allelic differences between parental populations for markers.
• Statistical analysis tools: Implements permutation and parametric procedures to generate neutral expectations, provides maximum likelihood estimates of hybrid indices from genotypic data, and includes graphical analyses for visualization of results.
• Implementation: Implemented in the R programming language.

Scientific Applications:

• Evolutionary biology and ecology: Analyzes introgression patterns to investigate the genetic architecture of reproductive isolation, adaptive introgression, and processes driving biological diversity.
• Empirical hybrid-zone studies: Applied to hybrid zones including Mus musculus and M. domesticus, and Helianthus petiolaris and H. annuus.

Methodology:

Uses multinomial regression to estimate genomic clines; applies permutation and parametric procedures to generate neutral expectations; uses maximum likelihood to estimate hybrid indices; performs simulation-based power analyses; implemented in R.