Theoretical and experimental assessment of genome-based prediction in landraces of allogamous crops.

Discovery and enrichment of favorable alleles in landraces are key to making them accessible for crop improvement. Here, we present two fundamentally different concepts for genome-based selection in landrace-derived maize populations, one based on doubled-haploid (DH) lines derived directly from individual landrace plants and the other based on crossing landrace plants to a capture line. For both types of populations, we show theoretically how allele frequencies of the ancestral landrace and the capture line translate into expectations for molecular and genetic variances. We show that the DH approach has clear advantages over gamete capture with generally higher prediction accuracies and no risk of masking valuable variation of the landrace. Prediction accuracies as high as 0.58 for dry matter yield in the DH population indicate high potential of genome-based selection. Based on a comparison among traits, we show that the genetic makeup of the capture line has great influence on the success of genome-based selection and that confounding effects between the alleles of the landrace and the capture line are best controlled for traits for which the capture line does not outperform the ancestral population per se or in testcrosses. Our results will guide the optimization of genome-enabled prebreeding schemes.

Accounting for epistasis improves genomic prediction of phenotypes with univariate and bivariate models across environments.

KEY MESSAGE: The accuracy of genomic prediction of phenotypes can be increased by including the top-ranked pairwise SNP interactions into the prediction model. We compared the predictive ability of various prediction models for a maize dataset derived from 910 doubled haploid lines from two European landraces (Kemater Landmais Gelb and Petkuser Ferdinand Rot), which were tested at six locations in Germany and Spain. The compared models were Genomic Best Linear Unbiased Prediction (GBLUP) as an additive model, Epistatic Random Regression BLUP (ERRBLUP) accounting for all pairwise SNP interactions, and selective Epistatic Random Regression BLUP (sERRBLUP) accounting for a selected subset of pairwise SNP interactions. These models have been compared in both univariate and bivariate statistical settings for predictions within and across environments. Our results indicate that modeling all pairwise SNP interactions into the univariate/bivariate model (ERRBLUP) is not superior in predictive ability to the respective additive model (GBLUP). However, incorporating only a selected subset of interactions with the highest effect variances in univariate/bivariate sERRBLUP can increase predictive ability significantly compared to the univariate/bivariate GBLUP. Overall, bivariate models consistently outperform univariate models in predictive ability. Across all studied traits, locations and landraces, the increase in prediction accuracy from univariate GBLUP to univariate sERRBLUP ranged from 5.9 to 112.4 percent, with an average increase of 47 percent. For bivariate models, the change ranged from -0.3 to + 27.9 percent comparing the bivariate sERRBLUP to the bivariate GBLUP, with an average increase of 11 percent. This considerable increase in predictive ability achieved by sERRBLUP may be of interest for "sparse testing" approaches in which only a subset of the lines/hybrids of interest is observed at each location.

European maize landraces made accessible for plant breeding and genome-based studies.

KEY MESSAGE: Doubled-haploid libraries from landraces capture native genetic diversity for a multitude of quantitative traits and make it accessible for breeding and genome-based studies. Maize landraces comprise large allelic diversity. We created doubled-haploid (DH) libraries from three European flint maize landraces and characterized them with respect to their molecular diversity, population structure, trait means, variances, and trait correlations. In total, 899 DH lines were evaluated using high-quality genotypic and multi-environment phenotypic data from up to 11 environments. The DH lines covered 95% of the molecular variation present in 35 landraces of an earlier study and represent the original three landrace populations in an unbiased manner. A comprehensive analysis of the target trait plant development at early growth stages as well as other important agronomic traits revealed large genetic variation for line per se and testcross performance. The majority of the 378 DH lines evaluated as testcrosses outperformed the commercial hybrids for early development. For total biomass yield, we observed a yield gap of 15% between mean testcross yield of the commercial hybrids and mean testcross yield of the DH lines. The DH lines also exhibited genetic variation for undesirable traits like root lodging and tillering, but correlations with target traits early development and yield were low or nonsignificant. The presented diversity atlas is a valuable, publicly available resource for genome-based studies to identify novel trait variation and evaluate the prospects of genomic prediction in landrace-derived material.

Genomic prediction within and across maize landrace derived populations using haplotypes.

Genomic prediction (GP) using haplotypes is considered advantageous compared to GP solely reliant on single nucleotide polymorphisms (SNPs), owing to haplotypes' enhanced ability to capture ancestral information and their higher linkage disequilibrium with quantitative trait loci (QTL). Many empirical studies supported the advantages of haplotype-based GP over SNP-based approaches. Nevertheless, the performance of haplotype-based GP can vary significantly depending on multiple factors, including the traits being studied, the genetic structure of the population under investigation, and the particular method employed for haplotype construction. In this study, we compared haplotype and SNP based prediction accuracies in four populations derived from European maize landraces. Populations comprised either doubled haploid lines (DH) derived directly from landraces, or gamete capture lines (GC) derived from crosses of the landraces with an inbred line. For two different landraces, both types of populations were generated, genotyped with 600k SNPs and phenotyped as lines per se for five traits. Our study explores three prediction scenarios: (i) within each of the four populations, (ii) across DH and GC populations from the same landrace, and (iii) across landraces using either DH or GC populations. Three haplotype construction methods were evaluated: 1. fixed-window blocks (FixedHB), 2. LD-based blocks (HaploView), and 3. IBD-based blocks (HaploBlocker). In within population predictions, FixedHB and HaploView methods performed as well as or slightly better than SNPs for all traits. HaploBlocker improved accuracy for certain traits but exhibited inferior performance for others. In prediction across populations, the parameter setting from HaploBlocker which controls the construction of shared haplotypes between populations played a crucial role for obtaining optimal results. When predicting across landraces, accuracies were low for both, SNP and haplotype approaches, but for specific traits substantial improvement was observed with HaploBlocker. This study provides recommendations for optimal haplotype construction and identifies relevant parameters for constructing haplotypes in the context of genomic prediction.

Genomic prediction using information across years with epistatic models and dimension reduction via haplotype blocks.

The importance of accurate genomic prediction of phenotypes in plant breeding is undeniable, as higher prediction accuracy can increase selection responses. In this regard, epistasis models have shown to be capable of increasing the prediction accuracy while their high computational load is challenging. In this study, we investigated the predictive ability obtained in additive and epistasis models when utilizing haplotype blocks versus pruned sets of SNPs by including phenotypic information from the last growing season. This was done by considering a single biological trait in two growing seasons (2017 and 2018) as separate traits in a multi-trait model. Thus, bivariate variants of the Genomic Best Linear Unbiased Prediction (GBLUP) as an additive model, Epistatic Random Regression BLUP (ERRBLUP) and selective Epistatic Random Regression BLUP (sERRBLUP) as epistasis models were compared with respect to their prediction accuracies for the second year. The prediction accuracies of bivariate GBLUP, ERRBLUP and sERRBLUP were assessed with eight phenotypic traits for 471/402 doubled haploid lines in the European maize landrace Kemater Landmais Gelb/Petkuser Ferdinand Rot. The results indicate that the obtained prediction accuracies are similar when utilizing a pruned set of SNPs or haplotype blocks, while utilizing haplotype blocks reduces the computational load significantly compared to the pruned sets of SNPs. The number of interactions considered in the model was reduced from 323.5/456.4 million for the pruned SNP panel to 4.4/5.5 million in the haplotype block dataset for Kemater and Petkuser landraces, respectively. Since the computational load scales linearly with the number of parameters in the model, this leads to a reduction in computational time of 98.9% from 13.5 hours for the pruned set of markers to 9 minutes for the haplotype block dataset. We further investigated the impact of genomic correlation, phenotypic correlation and trait heritability as factors affecting the bivariate models' prediction accuracy, identifying the genomic correlation between years as the most influential one. As computational load is substantially reduced, while the accuracy of genomic prediction is unchanged, the here proposed framework to use haplotype blocks in sERRBLUP provided a solution for the practical implementation of sERRBLUP in real breeding programs. Furthermore, our results indicate that sERRBLUP is not only suitable for prediction across different locations, but also for the prediction across growing seasons.

Genetic diversity of European maize landraces: Dataset on the molecular and phenotypic variation of derived doubled-haploid populations.

Genetic variation is the basis of selection, evolution and breeding. Maize landraces represent a rich source of allelic diversity, but their efficient utilization in breeding and research has been hampered by their heterogeneous and heterozygous nature and insufficient information about most accessions. While molecular inventories of germplasm repositories are growing steadily, linking these data to meaningful phenotypes for quantitative traits is challenging. Here, we present comprehensive molecular and phenotypic data for ∼1,000 doubled-haploid (DH) lines derived from three pre-selected European maize landraces. Due to their full homozygosity, the DH lines can be multiplied ad libitum and represent a powerful biological resource available to the community. The DH lines allow high-precision phenotyping in repeated experiments and reveal the full additive genetic variance of the population. The DH lines were evaluated for nine agronomically important, quantitative traits in multi-environment field trials comprising seven locations and two years. The DH populations revealed high genetic variance and high heritability for the analysed traits. The DH lines were genotyped with 600k SNP markers. After stringent quality filtering 500k markers remained for further analyses. This is the largest resource of landrace derived DH material in maize, unprecedented in its structure and dimension. The presented data are ideal for linking molecular variation to meaningful phenotypes. They can be used for genome-wide association studies, genomic prediction, and population genetic analyses as well as for developing and testing statistical methods. All plant material is available to the community for conducting additional experiments, extending the panel of traits and environments, and for testing the landrace-derived lines in combination with other genetic material.

Discovery of beneficial haplotypes for complex traits in maize landraces.

Genetic variation is of crucial importance for crop improvement. Landraces are valuable sources of diversity, but for quantitative traits efficient strategies for their targeted utilization are lacking. Here, we map haplotype-trait associations at high resolution in ~1000 doubled-haploid lines derived from three maize landraces to make their native diversity for early development traits accessible for elite germplasm improvement. A comparative genomic analysis of the discovered haplotypes in the landrace-derived lines and a panel of 65 breeding lines, both genotyped with 600k SNPs, points to untapped beneficial variation for target traits in the landraces. The superior phenotypic performance of lines carrying favorable landrace haplotypes as compared to breeding lines with alternative haplotypes confirms these findings. Stability of haplotype effects across populations and environments as well as their limited effects on undesired traits indicate that our strategy has high potential for harnessing beneficial haplotype variation for quantitative traits from genetic resources.