Exploration of selective genotyping and selective phenotyping for optimization of soybean genomic prediction models

Abstract

The accuracy of genomic selection can be affected by several factors including trait architecture and heritability, marker density, linkage disequilibrium between markers and trait loci, statistical models, training population size, composition, and population structure. The selection of a minimal and optimal marker set with high prediction accuracy as an alternative to reduce genotyping costs, computational time, and multicollinearity for the genomic selection is a challenging task. Furthermore, optimal training population size is mostly determined empirically, by random sampling a whole set of genotypes, which may not reflect the true relationships in the population and may lead to the loss of rare genotypes and alleles. Selective phenotyping could reduce the number of genotypes tested in the field while preserving the genetic diversity of the initial population. This study aimed to evaluate different methods of selective genotyping and phenotyping on the accuracy of genomic prediction for soybean yield.