Genomic Prediction of Hybrid Combinations in the Early Stages of a Maize Hybrid Breeding Pipeline

Abstract

Prediction of single-cross hybrid performance has been a major goal of plant breeders since the beginning of hybrid breeding. Genomic prediction has shown to be a promising approach, but only limited studies have examined the accuracy of predicting single cross performance. Most of the studies rather focused on predicting top cross performance using single tester to determine the inbred parent’s worth in hybrid combinations. Moreover, no studies have examined the potential of predicting single crosses made among random progenies derived from a series of biparental families, which resembles the structure of germplasm comprising the initial stages of a hybrid maize breeding pipeline. The main objective of this study was to evaluate the potential of genomic prediction for identifying superior single crosses early in the breeding pipeline and optimize its application. To accomplish these objectives, we designed and analyzed a novel population of single-cross hybrids representing the Iowa Stiff Stalk Synthetic/Non-Stiff Stalk heterotic pattern commonly used in the development of North American commercial maize hybrids. The single cross prediction accuracies estimated using cross-validation ranged from 0.40 to 0.74 for grain yield, 0.68 to 0.91 for plant height and 0.54 to 0.94 for staygreen depending on the number of tested parents of the single crosses. The genomic estimated general and specific combining abilities showed a clear advantage over the use of genomic covariances among single crosses, especially when one or both parents of the single cross were untested in hybrid combinations. Overall, our results suggest that genomic prediction of the performance of single crosses made using random progenies from the early stages of the breeding pipeline holds great potential to re-design hybrid breeding and increase its efficiency.

Beyond Genomic Prediction: Combining Different Types of omics Data Can Improve Prediction of Hybrid Performance in Maize

“…These small SCA variances explain why incorporating SCA effects into the prediction models did not further improve predictive abilities (Table S1 in File S2), similar to what was observed by Westhues et al (2017) in a related dataset for silage maize. In contrast, distinctly larger ratios of SCA to GCA variance were reported by Bernardo (1996) and Kadam et al (2016), where the latter study comprised materials originating exclusively from several North American Dent heterotic groups. Heritabilites on an entry-mean basis corresponded well to estimates published by Massman et al (2013), who reported 0.85 for GY and 0.98 for grain moisture among maize singlecrosses.…”

“…2013; Technow et al 2014;Kadam et al 2016). Traditionally, pedigree and genomic information on the parent lines have been used for predictions of breeding values, and the majority of studies in the past decade focused on conceiving and improving algorithms for exploiting the full potential of these data (Meuwissen et al 2001;Maenhout et al 2010;Habier et al 2011).…”

Editorial: Crop Yield and Quality Response to the Interaction Between Environment and Genetic Factors

“…Because of the complexities of the environment, which include climate, soil condition, and crop management practices, one of the most difficult aspects of predicting crop yield is applying appropriate models including climatic and phenotypic data and molecular markers (Kadam et al, 2016). Climatic and phenotypic data can be easily recorded with automatic instruments; however, they often fluctuate significantly.…”

Response Surface Analysis of Genomic Prediction Accuracy Values Using Quality Control Covariates in Soybean

“…Thus, there is not a conventional value used in all species. For example, in maize 21 and wheat, 22 a MAF of 0.05 was used for GBS and an Infinitum SNP array, respectively, while 23 considered a cut-off of 0.01 in wheat for GBS data. PMMS is the percentage of missing marker scores is the percentage of marker scores in the genomic data set that are missing.…”