KinInfor

KinInfor evaluates the informativeness of genetic markers to quantify their utility for inferring pairwise relationships (R) and estimating pairwise relatedness (r) using information-theoretic metrics.

Key Features:

• I(R) - Informativeness for Relationship: Quantifies the amount of information a marker provides for inferring pairwise relationships (R).
• I(r) - Informativeness for Relatedness: Measures the informativeness of markers for estimating pairwise relatedness (r).
• PW(R) - Statistical Power Calculation: Computes the power of a set of markers (PW(R)) to distinguish between two candidate relationships using a fast algorithm.
• RMSD - Reciprocal Mean Squared Deviations: Reports the reciprocal of the mean squared deviations of relatedness estimates to assess estimator performance.
• Marker type support: Applies to both dominant and codominant genetic markers.
• Ploidy support: Handles haploid and diploid individuals.
• Error and mutation modeling: Accounts for potential mutations and typing errors in genotype data.
• Implementation: Implemented as a Fortran program based on information-theoretic principles.
• Analytical and empirical evaluation: Statistical properties of I(R), I(r), PW(R), and RMSD have been investigated analytically and examined on simulated and empirical datasets.

Scientific Applications:

• Population genetics: Selects and ranks markers for relationship and relatedness inference in population-genetic analyses.
• Conservation biology: Optimizes marker panels to assess kinship and relatedness for conservation management.
• Evolutionary studies: Supports inference of genetic relationships and relatedness in evolutionary research.
• Marker panel design: Aids optimization of marker selection to maximize statistical power while minimizing genotyping effort.

Methodology:

Computes I(R), I(r), PW(R), and RMSD using information-theoretic calculations and a fast algorithm for PW(R); applies to dominant and codominant markers and to haploid and diploid individuals while accounting for mutations and typing errors; statistical properties were investigated analytically and validated on simulated and empirical datasets in a Fortran implementation.