Efficient computation of the genomic relationship matrix and other matrices used in single‐step evaluation

Ignácio Aguilar, I. Misztal, Andrés Legarra, S. Tsuruta

Journal of Animal Breeding and Genetics · 2011 · 175 citations · 20 references

Concepts

TL;DR

Genomic evaluations combine phenotypic, pedigree, and genomic data, requiring the inverse of the relationship matrix. The study aims to identify efficient computational options for constructing and inverting relationship matrices from genomic markers and pedigree information. Simulating 40 000 SNPs for up to 30 000 animals, the authors computed genomic relationship matrices using simple loops, optimized loops, and a matrix multiplication subroutine, and inverted them with a generalized inverse algorithm and LAPACK, applying further optimizations via code modifications or library calls. With the most efficient choices and parallel processing, matrices for 30 000 animals can be generated in a few hours, proving that the required matrices for a unified approach can be computed efficiently.

Abstract

Summary Genomic evaluations can be calculated using a unified procedure that combines phenotypic, pedigree and genomic information. Implementation of such a procedure requires the inverse of the relationship matrix based on pedigree and genomic relationships. The objective of this study was to investigate efficient computing options to create relationship matrices based on genomic markers and pedigree information as well as their inverses. SNP maker information was simulated for a panel of 40 K SNPs, with the number of genotyped animals up to 30 000. Matrix multiplication in the computation of the genomic relationship was by a simple ‘do’ loop, by two optimized versions of the loop, and by a specific matrix multiplication subroutine. Inversion was by a generalized inverse algorithm and by a LAPACK subroutine. With the most efficient choices and parallel processing, creation of matrices for 30 000 animals would take a few hours. Matrices required to implement a unified approach can be computed efficiently. Optimizations can be either by modifications of existing code or by the use of efficient automatic optimizations provided by open source or third‐party libraries.

References

20