Forecasting the accuracy of genomic prediction with different selection targets in the training and prediction set as well as truncation selection

Unraveling the potential of phenomic selection within and among diverse breeding material of maize (Zea mays L.)

Genomic selection is a well-investigated approach that facilitates and supports selection decisions for complex traits and has meanwhile become a standard tool in modern plant breeding. Phenomic selection has only recently been suggested and uses the same statistical procedures to predict the targeted traits but replaces marker data with near-infrared spectroscopy data. It may represent an attractive low-cost, high-throughput alternative but has not been sufficiently studied until now. Here, we used 400 genotypes of maize (Zea mays L.) comprising elite lines of the Flint and Dent heterotic pools as well as 6 Flint landraces, which were phenotyped in multienvironment trials for anthesis-silking-interval, early vigor, final plant height, grain dry matter content, grain yield, and phosphorus concentration in the maize kernels, to compare the predictive abilities of genomic as well as phenomic prediction under different scenarios. We found that both approaches generally achieved comparable predictive abilities within material groups. However, phenomic prediction was less affected by population structure and performed better than its genomic counterpart for predictions among diverse groups of breeding material. We therefore conclude that phenomic prediction is a promising tool for practical breeding, for instance when working with unknown and rather diverse germplasm. Moreover, it may make the highly monopolized sector of plant breeding more accessible also for low-tech institutions by combining well established, widely available, and cost-efficient spectral phenotyping with the statistical procedures elaborated for genomic prediction - while achieving similar or even better results than with marker data.

Calibration and validation of predicted genomic breeding values in an advanced cycle maize population

Model training on data from all selection cycles yielded the highest prediction accuracy by attenuating specific effects of individual cycles. Expected reliability was a robust predictor of accuracies obtained with different calibration sets. The transition from phenotypic to genome-based selection requires a profound understanding of factors that determine genomic prediction accuracy. We analysed experimental data from a commercial maize breeding programme to investigate if genomic measures can assist in identifying optimal calibration sets for model training. The data set consisted of six contiguous selection cycles comprising testcrosses of 5968 doubled haploid lines genotyped with a minimum of 12,000 SNP markers. We evaluated genomic prediction accuracies in two independent prediction sets in combination with calibration sets differing in sample size and genomic measures (effective sample size, average maximum kinship, expected reliability, number of common polymorphic SNPs and linkage phase similarity). Our results indicate that across selection cycles prediction accuracies were as high as 0.57 for grain dry matter yield and 0.76 for grain dry matter content. Including data from all selection cycles in model training yielded the best results because interactions between calibration and prediction sets as well as the effects of different testers and specific years were attenuated. Among genomic measures, the expected reliability of genomic breeding values was the best predictor of empirical accuracies obtained with different calibration sets. For grain yield, a large difference between expected and empirical reliability was observed in one prediction set. We propose to use this difference as guidance for determining the weight phenotypic data of a given selection cycle should receive in model retraining and for selection when both genomic breeding values and phenotypes are available.

Back to the future: Implications of genetic complexity for the structure of hybrid breeding programs

Commercial hybrid breeding operations can be described as decentralized networks of smaller, more or less isolated breeding programs. There is further a tendency for the disproportionate use of successful inbred lines for generating the next generation of recombinants, which has led to a series of significant bottlenecks, particularly in the history of the North American and European maize germplasm. Both the decentralization and the disproportionate contribution of inbred lines reduce effective population size and constrain the accessible genetic space. Under these conditions, long-term response to selection is not expected to be optimal under the classical infinitesimal model of quantitative genetics. In this study, we therefore aim to propose a rationale for the success of large breeding operations in the context of genetic complexity arising from the structure and properties of interactive genetic networks. For this, we use simulations based on the NK model of genetic architecture. We indeed found that constraining genetic space through program decentralization and disproportionate contribution of parental inbred lines, is required to expose additive genetic variation and thus facilitate heritable genetic gains under high levels of genetic complexity. These results introduce new insights into why the historically grown structure of hybrid breeding programs was successful in improving the yield potential of hybrid crops over the last century. We also hope that a renewed appreciation for “why things worked” in the past can guide the adoption of novel technologies and the design of future breeding strategies for navigating biological complexity.

Use of F2 Bulks in Training Sets for Genomic Prediction of Combining Ability and Hybrid Performance

Developing training sets for genomic prediction in hybrid crops requires producing hybrid seed for a large number of entries. In autogamous crop species (e.g., wheat, rice, rapeseed, cotton) this requires elaborate hybridization systems to prevent self-pollination and presents a significant impediment to the implementation of hybrid breeding in general and genomic selection in particular. An alternative to F1 hybrids are bulks of F2 seed from selfed F1 plants (F1:2). Seed production for F1:2 bulks requires no hybridization system because the number of F1 plants needed for producing enough F1:2 seed for multi-environment testing can be generated by hand-pollination. This study evaluated the suitability of F1:2 bulks for use in training sets for genomic prediction of F1 level general combining ability and hybrid performance, under different degrees of divergence between heterotic groups and modes of gene action, using quantitative genetic theory and simulation of a genomic prediction experiment. The simulation, backed by theory, showed that F1:2 training sets are expected to have a lower prediction accuracy relative to F1 training sets, particularly when heterotic groups have strongly diverged. The accuracy penalty, however, was only modest and mostly because of a lower heritability, rather than because of a difference in F1 and F1:2 genetic values. It is concluded that resorting to F1:2 bulks is, in theory at least, a promising approach to remove the significant complication of a hybridization system from the breeding process.

Across-years prediction of hybrid performance in maize using genomics

Inclusion of historical training data improved the genomics-based prediction of performance of maize hybrids, the extent depending on the phenotypic trait and genotype-by-year interaction. Prediction of hybrid performance using existing phenotypic data on previous hybrids combined with molecular data collected on the parent lines allows to identify the most promising candidates from a huge number of possible hybrids at an early stage. Phenotypic data on yield and dry matter of 1970 grain maize hybrids from 19 years of a public breeding program were aggregated considering the underlying structure of factorial sets of hybrids. Pedigree records and 50 K SNP data were collected on their 170 Dent and 127 Flint parent lines. The performance of untested hybrids was predicted by best linear unbiased predictors (BLUP) on basis of pedigree or genomic data. For composition of training sets (TRN) and test sets (TST), three schemes for collecting factorials from specific years were employed which resulted in 490 scenarios. For each scenario, the predictive ability and genomic relationship between TRN and TST hybrids were determined. For extended TRNs, where earlier years were successively added to the TRN, the maximum relationship increased and the predictive ability improved, with the extent of the latter depending on the phenotypic trait and its genotype-by-year interaction. Genomic BLUP outperformed pedigree BLUP and better utilized the early years' data, especially for prediction of hybrids from factorials in a more distant future. This study on hybrid prediction in grain maize illustrated that including historical phenotypic data for training, although consisting of less related genotypes, can improve genomic prediction and enables optimization of hybrid variety development.

Enhancing genetic gain in the era of molecular breeding

As one of the important concepts in conventional quantitative genetics and breeding, genetic gain can be defined as the amount of increase in performance that is achieved annually through artificial selection. To develop pro ducts that meet the increasing demand of mankind, especially for food and feed, in addition to various industrial uses, breeders are challenged to enhance the potential of genetic gain continuously, at ever higher rates, while they close the gaps that remain between the yield potential in breeders' demonstration trials and the actual yield in farmers' fields. Factors affecting genetic gain include genetic variation available in breeding materials, heritability for traits of interest, selection intensity, and the time required to complete a breeding cycle. Genetic gain can be improved through enhancing the potential and closing the gaps, which has been evolving and complemented with modern breeding techniques and platforms, mainly driven by molecular and genomic tools, combined with improved agronomic practice. Several key strategies are reviewed in this article. Favorable genetic variation can be unlocked and created through molecular and genomic approaches including mutation, gene mapping and discovery, and transgene and genome editing. Estimation of heritability can be improved by refining field experiments through well-controlled and precisely assayed environmental factors or envirotyping, particularly for understanding and controlling spatial heterogeneity at the field level. Selection intensity can be significantly heightened through improvements in the scale and precision of genotyping and phenotyping. The breeding cycle time can be shortened by accelerating breeding procedures through integrated breeding approaches such as marker-assisted selection and doubled haploid development. All the strategies can be integrated with other widely used conventional approaches in breeding programs to enhance genetic gain. More transdisciplinary approaches, team breeding, will be required to address the challenge of maintaining a plentiful and safe food supply for future generations. New opportunities for enhancing genetic gain, a high efficiency breeding pipeline, and broad-sense genetic gain are also discussed prospectively.

Accuracy of Genomic Prediction in Synthetic Populations Depending on the Number of Parents, Relatedness, and Ancestral Linkage Disequilibrium

Synthetics play an important role in quantitative genetic research and plant breeding, but few studies have investigated the application of genomic prediction (GP) to these populations. Synthetics are generated by intermating a small number of parents ([Formula: see text] and thereby possess unique genetic properties, which make them especially suited for systematic investigations of factors contributing to the accuracy of GP. We generated synthetics in silico from [Formula: see text]2 to 32 maize (Zea mays L.) lines taken from an ancestral population with either short- or long-range linkage disequilibrium (LD). In eight scenarios differing in relatedness of the training and prediction sets and in the types of data used to calculate the relationship matrix (QTL, SNPs, tag markers, and pedigree), we investigated the prediction accuracy (PA) of Genomic best linear unbiased prediction (GBLUP) and analyzed contributions from pedigree relationships captured by SNP markers, as well as from cosegregation and ancestral LD between QTL and SNPs. The effects of training set size [Formula: see text] and marker density were also studied. Sampling few parents ([Formula: see text]) generates substantial sample LD that carries over into synthetics through cosegregation of alleles at linked loci. For fixed [Formula: see text], [Formula: see text] influences PA most strongly. If the training and prediction set are related, using [Formula: see text] parents yields high PA regardless of ancestral LD because SNPs capture pedigree relationships and Mendelian sampling through cosegregation. As [Formula: see text] increases, ancestral LD contributes more information, while other factors contribute less due to lower frequencies of closely related individuals. For unrelated prediction sets, only ancestral LD contributes information and accuracies were poor and highly variable for [Formula: see text] due to large sample LD. For large [Formula: see text], achieving moderate accuracy requires large [Formula: see text], long-range ancestral LD, and high marker density. Our approach for analyzing PA in synthetics provides new insights into the prospects of GP for many types of source populations encountered in plant breeding.