Computing inbreeding coefficients accounting for unknown parents using pedigree-based metafounders.

Area of Science:

• Animal Genetics
• Quantitative Genetics
• Bioinformatics

Background:

• Missing pedigree data in livestock populations leads to inaccurate numerator relationship and inbreeding coefficients.
• Existing recursive algorithms for estimating inbreeding with unknown parents are computationally intensive for deep pedigrees.

Purpose of the Study:

• To develop and evaluate a novel pedigree-based metafounder (MF) approach for calculating nonzero inbreeding coefficients.
• To assess the computational efficiency and accuracy of the MF approach compared to recursive algorithms.

Main Methods:

• Unknown parents were assigned to metafounders (MF) based on offspring birth year.
• MF self-relationships were estimated using average inbreeding of known animals.
• Inbreeding coefficients were computed using a modified Sargolzaei and Iwaisaki (2004) algorithm with MF, and the MF relationship matrix was iteratively constructed.

Main Results:

• The MF approach significantly reduced computation time (70s vs. 300s per iteration) for a large Japanese Holstein pedigree.
• The MF approach demonstrated high accuracy, with the 'maximum' assumption for MF relationships yielding the best correlation between estimated and true inbreeding.
• Both MF and recursive algorithms converged to identical results when the 'maximum' assumption was used for MF relationships.

Conclusions:

• The pedigree-based metafounder (MF) approach provides a computationally efficient and accurate method for estimating inbreeding coefficients in populations with missing pedigree information.
• The MF approach facilitates rapid derivation of the numerator relationship matrix subsets related to unknown parents.
• The choice of assumption for relationships between metafounders impacts the accuracy of calculated inbreeding and relationship coefficients.