CoCoa: a software tool for estimating the coefficient of coancestry from multilocus genotype data

Area of Science:

• Quantitative genetics
• Bioinformatics
• Plant and animal breeding

Background:

• Linear mixed models are crucial for analyzing phenotypic data in breeding programs and association studies.
• Estimating the genetic covariance matrix using marker-based coancestry methods can introduce bias, especially in breeding populations.
• Accurate coancestry estimation is vital for reliable genetic analyses.

Purpose of the Study:

• To introduce CoCoa, a software package for estimating marker-based coancestry matrices.
• To provide tools for selecting the best covariance structure for phenotypic data analysis.
• To implement advanced methods like the Weighted Alikeness in State (WAIS) estimator for hybrid breeding programs.

Main Methods:

• CoCoa implements various commonly used marker-based coancestry estimation procedures.
• The software includes two bending heuristics for inverting ill-conditioned coancestry matrices.
• It offers matrix manipulation tools and the WAIS estimator.

Main Results:

• CoCoa facilitates the selection of optimal covariance structures, leading to improved model fit.
• This improved fit enhances statistical power and type I error control in association studies.
• Accuracy in phenotypic prediction studies is also improved.

Conclusions:

• CoCoa provides a robust solution for coancestry estimation in breeding programs.
• The software enhances the reliability and accuracy of genetic analyses.
• CoCoa is freely available with comprehensive documentation and system support.