Genome-based prediction of test cross performance in two subsequent breeding cycles

Hofheinz, Nina; Borchardt, Dietrich; Weissleder, Knuth; Frisch, Matthias

doi:10.1007/s00122-012-1940-5

Cited by 65 publications

(55 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We hypothesize that in experimental data, a certain level of predictive ability was due to the presence of genetic relationships but this was not the case in computer simulations. With experimental data, all four methods seemed to exploit genetic relationships for prediction by assigning equal genotypic values to relatives sharing a large fraction of the genome (Hofheinz et al 2012). Thus, for prediction of genomic breeding values, RR-BLUP is a good choice because this method is robust and computationally efficient.…”

Section: Discrepancies In Methods Comparisons Between Simulated and Exmentioning

confidence: 99%

Genome-Wide Prediction of Traits with Different Genetic Architecture Through Efficient Variable Selection

Wimmer

Lehermeier

Albrecht³

et al. 2013

Genetics

122

115

View full text Add to dashboard Cite

In genome-based prediction there is considerable uncertainty about the statistical model and method required to maximize prediction accuracy. For traits influenced by a small number of quantitative trait loci (QTL), predictions are expected to benefit from methods performing variable selection [e.g., BayesB or the least absolute shrinkage and selection operator (LASSO)] compared to methods distributing effects across the genome [ridge regression best linear unbiased prediction (RR-BLUP)]. We investigate the assumptions underlying successful variable selection by combining computer simulations with large-scale experimental data sets from rice (Oryza sativa L.), wheat (Triticum aestivum L.), and Arabidopsis thaliana (L.). We demonstrate that variable selection can be successful when the number of phenotyped individuals is much larger than the number of causal mutations contributing to the trait. We show that the sample size required for efficient variable selection increases dramatically with decreasing trait heritabilities and increasing extent of linkage disequilibrium (LD). We contrast and discuss contradictory results from simulation and experimental studies with respect to superiority of variable selection methods over RR-BLUP. Our results demonstrate that due to long-range LD, medium heritabilities, and small sample sizes, superiority of variable selection methods cannot be expected in plant breeding populations even for traits like FRIGIDA gene expression in Arabidopsis and flowering time in rice, assumed to be influenced by a few major QTL. We extend our conclusions to the analysis of whole-genome sequence data and infer upper bounds for the number of causal mutations which can be identified by LASSO. Our results have major impact on the choice of statistical method needed to make credible inferences about genetic architecture and prediction accuracy of complex traits. G ENOME-BASED prediction of genotypic values from marker or sequence information has been shown to be a valuable tool in plant and animal breeding (Meuwissen et al. 2001;Meuwissen and Goddard 2010;Albrecht et al. 2011). A series of statistical methods mainly differing in the extent of regularization and variable selection has been proposed in the literature (a review is in de los Campos et al. 2013). Simulation studies revealed clear differences between methods with respect to their predictive ability. Several factors affecting the prediction performance of these methods such as genetic trait architecture, span of linkage disequilibrium (LD), sample size, trait heritability, and marker density have been identified (Zhong et al. 2009;Daetwyler et al. 2010;Habier et al. 2010). However, how these methods account for the respective factors is still not fully understood, causing uncertainty about the best choice of method for a given population and trait.High-throughput genotyping platforms deliver data sets where the number of available observations n is typically smaller than the number of markers p. In these high-dimensional data sets...

show abstract

Section: Discrepancies In Methods Comparisons Between Simulated and Exmentioning

confidence: 99%

Genome-Wide Prediction of Traits with Different Genetic Architecture Through Efficient Variable Selection

Wimmer

Lehermeier

Albrecht³

et al. 2013

Genetics

122

115

View full text Add to dashboard Cite

show abstract

“…By using ridge regression best linear unbiased prediction (RR-BLUP) [29], the estimated marker effects ( ) were estimated based on a mixed model equation, , where is the transpose of 1 N , X T represents the transpose of X , I is an identity matrix, λ represents a penalty parameter, and denotes the estimated overall mean. The penalty parameter can be calculated as , where m is the number of markers and h 2 refers to the heritability of the estimation set [30]. Then the genetic values were predicted as , where is the estimated marker effect.…”

Section: Methodsmentioning

confidence: 99%

The accuracy of prediction of genomic selection in elite hybrid rye populations surpasses the accuracy of marker-assisted selection and is equally augmented by multiple field evaluation locations and test years

et al. 2014

View full text Add to dashboard Cite

BackgroundMarker-assisted selection (MAS) and genomic selection (GS) based on genome-wide marker data provide powerful tools to predict the genotypic value of selection material in plant breeding. However, case-to-case optimization of these approaches is required to achieve maximum accuracy of prediction with reasonable input.ResultsBased on extended field evaluation data for grain yield, plant height, starch content and total pentosan content of elite hybrid rye derived from testcrosses involving two bi-parental populations that were genotyped with 1048 molecular markers, we compared the accuracy of prediction of MAS and GS in a cross-validation approach. MAS delivered generally lower and in addition potentially over-estimated accuracies of prediction than GS by ridge regression best linear unbiased prediction (RR-BLUP). The grade of relatedness of the plant material included in the estimation and test sets clearly affected the accuracy of prediction of GS. Within each of the two bi-parental populations, accuracies differed depending on the relatedness of the respective parental lines. Across populations, accuracy increased when both populations contributed to estimation and test set. In contrast, accuracy of prediction based on an estimation set from one population to a test set from the other population was low despite that the two bi-parental segregating populations under scrutiny shared one parental line. Limiting the number of locations or years in field testing reduced the accuracy of prediction of GS equally, supporting the view that to establish robust GS calibration models a sufficient number of test locations is of similar importance as extended testing for more than one year.ConclusionsIn hybrid rye, genomic selection is superior to marker-assisted selection. However, it achieves high accuracies of prediction only for selection candidates closely related to the plant material evaluated in field trials, resulting in a rather pessimistic prognosis for distantly related material. Both, the numbers of evaluation locations and testing years in trials contribute equally to prediction accuracy.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2164-15-556) contains supplementary material, which is available to authorized users.

show abstract

“…A small number of studies discuss the implementation of GS in other (non-woody) crop species including brassica (Cowling & Balázs, 2010;Cowling et al, 2009;Cullis, Smith, Beeck, & Cowling, 2010;Würschum, Abel, & Zhao, 2014), oats (Asoro et al, 2011), potato (Barrell, Meiyalaghan, Jacobs, & Conner, 2013), sugar beet (Hofheinz, Borchardt, Weissleder, & Frisch, 2012;Würschum, Reif, Kraft, Janssen, & Zhao, 2013), sugarcane (Gouy et al, 2013) or soybean (Shu et al, 2013).…”

Section: Projections In Other Cropsmentioning

confidence: 99%

“…Estimates from related lines of the same breeding cycle will lead to considerable gain in response to selection while estimates from a previous breeding cycle might only be accurate predictors for highly heritable traits (Hofheinz et al, 2012). Training in a diversity population had been suggested, but the genetic distance to the selection candidates as well as the influence of GxE need to be investigated (Hofheinz et al, 2012;Würschum et al, 2013).…”

Section: Projections In Other Cropsmentioning

confidence: 99%

Goals and hurdles for a successful implementation of genomic selection in breeding programme for selected annual and perennial crops

Jonas

Koning

2016

Biotechnology and Genetic Engineering Reviews

View full text Add to dashboard Cite

Genomic Selection is an important topic in quantitative genetics and breeding. Not only does it allow the full use of current molecular genetic technologies, it stimulates also the development of new methods and models. Genomic selection, if fully implemented in commercial farming, should have a major impact on the productivity of various agricultural systems. But suggested approaches need to be applicable in commercial breeding populations. Many of the published research studies focus on methodologies. We conclude from the reviewed publications, that a stronger focus on strategies for the implementation of genomic selection in advanced breeding lines, introduction of new varieties, hybrids or multi-line crosses is needed. Efforts to find solutions for a better prediction and integration of environmental influences need to continue within applied breeding schemes. Goals of the implementation of genomic selection into crop breeding should be carefully defined and crop breeders in the private sector will play a substantial part in the decision-making process. However, the lack of published results from studies within, or in collaboration with, private companies diminishes the knowledge on the status of genomic selection within applied breeding programmes. Studies on the implementation of genomic selection in plant breeding need to evaluate models and methods with an enhanced emphasis on population-specific requirements and production environments. Adaptation of methods to breeding schemes or changes to breeding programmes for a better integration of genomic selection strategies are needed across species. More openness with a continuous exchange will contribute to successes.

show abstract

Genome-based prediction of test cross performance in two subsequent breeding cycles

Cited by 65 publications

References 18 publications

Genome-Wide Prediction of Traits with Different Genetic Architecture Through Efficient Variable Selection

Genome-Wide Prediction of Traits with Different Genetic Architecture Through Efficient Variable Selection

The accuracy of prediction of genomic selection in elite hybrid rye populations surpasses the accuracy of marker-assisted selection and is equally augmented by multiple field evaluation locations and test years

Goals and hurdles for a successful implementation of genomic selection in breeding programme for selected annual and perennial crops

Contact Info

Product

Resources

About