Ridge Regression Best Linear Unbiased Prediction Research Articles

Investigating genomic prediction strategies for grain carotenoid traits in a tropical/subtropical maize panel.

Vitamin A deficiency remains prevalent on a global scale, including in regions where maize constitutes a high percentage of human diets. One solution for alleviating this deficiency has been to increase grain concentrations of provitamin A carotenoids in maize (Zea mays ssp. mays L.)-an example of biofortification. The International Maize and Wheat Improvement Center (CIMMYT) developed a Carotenoid Association Mapping panel of 380 inbred lines adapted to tropical and subtropical environments that have varying grain concentrations of provitamin A and other health-beneficial carotenoids. Several major genes have been identified for these traits, 2 of which have particularly been leveraged in marker-assisted selection. This project assesses the predictive ability of several genomic prediction strategies for maize grain carotenoid traits within and between 4 environments in Mexico. Ridge Regression-Best Linear Unbiased Prediction, Elastic Net, and Reproducing Kernel Hilbert Spaces had high predictive abilities for all tested traits (β-carotene, β-cryptoxanthin, provitamin A, lutein, and zeaxanthin) and outperformed Least Absolute Shrinkage and Selection Operator. Furthermore, predictive abilities were higher when using genome-wide markers rather than only the markers proximal to 2 or 13 genes. These findings suggest that genomic prediction models using genome-wide markers (and assuming equal variance of marker effects) are worthwhile for these traits even though key genes have already been identified, especially if breeding for additional grain carotenoid traits alongside β-carotene. Predictive ability was maintained for all traits except lutein in between-environment prediction. The TASSEL (Trait Analysis by aSSociation, Evolution, and Linkage) Genomic Selection plugin performed as well as other more computationally intensive methods for within-environment prediction. The findings observed herein indicate the utility of genomic prediction methods for these traits and could inform their resource-efficient implementation in biofortification breeding programs.

Ridge regression and deep learning models for genome-wide selection of complex traits in New Mexican Chile peppers

BackgroundGenomewide prediction estimates the genomic breeding values of selection candidates which can be utilized for population improvement and cultivar development. Ridge regression and deep learning-based selection models were implemented for yield and agronomic traits of 204 chile pepper genotypes evaluated in multi-environment trials in New Mexico, USA.ResultsAccuracy of prediction differed across different models under ten-fold cross-validations, where high prediction accuracy was observed for highly heritable traits such as plant height and plant width. No model was superior across traits using 14,922 SNP markers for genomewide selection. Bayesian ridge regression had the highest average accuracy for first pod date (0.77) and total yield per plant (0.33). Multilayer perceptron (MLP) was the most superior for flowering time (0.76) and plant height (0.73), whereas the genomic BLUP model had the highest accuracy for plant width (0.62). Using a subset of 7,690 SNP loci resulting from grouping markers based on linkage disequilibrium coefficients resulted in improved accuracy for first pod date, ten pod weight, and total yield per plant, even under a relatively small training population size for MLP and random forest models. Genomic and ridge regression BLUP models were sufficient for optimal prediction accuracies for small training population size. Combining phenotypic selection and genomewide selection resulted in improved selection response for yield-related traits, indicating that integrated approaches can result in improved gains achieved through selection.ConclusionsAccuracy values for ridge regression and deep learning prediction models demonstrate the potential of implementing genomewide selection for genetic improvement in chile pepper breeding programs. Ultimately, a large training data is relevant for improved genomic selection accuracy for the deep learning models.

Using encrypted genotypes and phenotypes for collaborative genomic analyses to maintain data confidentiality.

To adhere to and capitalize on the benefits of the FAIR (findable, accessible, interoperable, and reusable) principles in agricultural genome-to-phenome studies, it is crucial to address privacy and intellectual property issues that prevent sharing and reuse of data in research and industry. Direct sharing of genotype and phenotype data is often prohibited due to intellectual property and privacy concerns. Thus, there is a pressing need for encryption methods that obscure confidential aspects of the data, without affecting the outcomes of certain statistical analyses. A homomorphic encryption method for genotypes and phenotypes (HEGP) has been proposed for single-marker regression in genome-wide association studies (GWAS) using linear mixed models with Gaussian errors. This methodology permits frequentist likelihood-based parameter estimation and inference. In this paper, we extend HEGP to broader applications in genome-to-phenome analyses. We show that HEGP is suited to commonly used linear mixed models for genetic analyses of quantitative traits including genomic best linear unbiased prediction (GBLUP) and ridge-regression best linear unbiased prediction (RR-BLUP), as well as Bayesian variable selection methods (e.g. those in Bayesian Alphabet), for genetic parameter estimation, genomic prediction, and GWAS. By advancing the capabilities of HEGP, we offer researchers and industry professionals a secure and efficient approach for collaborative genomic analyses while preserving data confidentiality.

Salinity stress tolerance prediction for biomass-related traits in maize (Zea mays L.) using genome-wide markers.

Maize (Zea mays L.) is the third most important cereal crop after rice (Oryza sativa) and wheat (Triticum aestivum). Salinity stress significantly affects vegetative biomass and grain yield and, therefore, reduces the food and silage productivity of maize. Selecting salt-tolerant genotypes is a cumbersome and time-consuming process that requires meticulous phenotyping. To predict salt tolerance in maize, we estimated breeding values for four biomass-related traits, including shoot length, shoot weight, root length, and root weight under salt-stressed and controlled conditions. A five-fold cross-validation method was used to select the best model among genomic best linear unbiased prediction (GBLUP), ridge-regression BLUP(rrBLUP), extendedGBLUP, Bayesian Lasso, Bayesian ridge regression, BayesA, BayesB, and BayesC. Examination of the effect of different marker densities on prediction accuracy revealed that a set of low-density single nucleotide polymorphisms obtained through filtering based on a combination of analysis of variance and linkage disequilibrium provided the best prediction accuracy for all the traits. The average prediction accuracy in cross-validations ranged from 0.46 to 0.77 across the four derived traits. The GBLUP, rrBLUP, and all Bayesian models except BayesB demonstrated comparable levels of prediction accuracy that were superior to the other modeling approaches. These findings provide a roadmap for the deployment and optimization of genomic selection in breeding for salt tolerance in maize.

Genome-wide association mapping and genomic prediction of agronomical traits and breeding values in Iranian wheat under rain-fed and well-watered conditions

BackgroundThe markers detected by genome-wide association study (GWAS) make it possible to dissect genetic structure and diversity at many loci. This can enable a wheat breeder to reveal and used genomic loci controlling drought tolerance. This study was focused on determining the population structure of Iranian 208 wheat landraces and 90 cultivars via genotyping-by-sequencing (GBS) and also on detecting marker-trait associations (MTAs) by GWAS and genomic prediction (GS) of wheat agronomic traits for drought-tolerance breeding. GWASs were conducted using both the original phenotypes (pGWAS) and estimated breeding values (eGWAS). The bayesian ridge regression (BRR), genomic best linear unbiased prediction (gBLUP), and ridge regression-best linear unbiased prediction (rrBLUP) approaches were used to estimate breeding values and estimate prediction accuracies in genomic selection.ResultsPopulation structure analysis using 2,174,975 SNPs revealed four genetically distinct sub-populations from wheat accessions. D-Genome harbored the lowest number of significant marker pairs and the highest linkage disequilibrium (LD), reflecting different evolutionary histories of wheat genomes. From pGWAS, BRR, gBLUP, and rrBLUP, 284, 363, 359 and 295 significant MTAs were found under normal and 195, 365, 362 and 302 under stress conditions, respectively. The gBLUP with the most similarity (80.98 and 71.28% in well-watered and rain-fed environments, correspondingly) with the pGWAS method in the terms of discovered significant SNPs, suggesting the potential of gBLUP in uncovering SNPs. Results from gene ontology revealed that 29 and 30 SNPs in the imputed dataset were located in protein-coding regions for well-watered and rain-fed conditions, respectively. gBLUP model revealed genetic effects better than other models, suggesting a suitable tool for genome selection in wheat.ConclusionWe illustrate that Iranian landraces of bread wheat contain novel alleles that are adaptive to drought stress environments. gBLUP model can be helpful for fine mapping and cloning of the relevant QTLs and genes, and for carrying out trait introgression and marker-assisted selection in both normal and drought environments in wheat collections.

Genome-Wide Association Study (GWAS) and genome prediction of seedling salt tolerance in bread wheat (Triticum aestivum L.)

BackgroundSalinity tolerance in wheat is imperative for improving crop genetic capacity in response to the expanding phenomenon of soil salinization. However, little is known about the genetic foundation underlying salinity tolerance at the seedling growth stage of wheat. Herein, a GWAS analysis was carried out by the random-SNP-effect mixed linear model (mrMLM) multi-locus model to uncover candidate genes responsible for salt tolerance at the seedling stage in 298 Iranian bread wheat accessions, including 208 landraces and 90 cultivars.ResultsA total of 29 functional marker-trait associations (MTAs) were detected under salinity, 100 mM NaCl (sodium chloride). Of these, seven single nucleotide polymorphisms (SNPs) including rs54146, rs257, rs37983, rs18682, rs55629, rs15183, and rs63185 with R2 ≥ 10% were found to be linked with relative water content, root fresh weight, root dry weight, root volume, shoot high, proline, and shoot potassium (K+), respectively. Further, a total of 27 candidate genes were functionally annotated to be involved in response to the saline environment. Most of these genes have key roles in photosynthesis, response to abscisic acid, cell redox homeostasis, sucrose and carbohydrate metabolism, ubiquitination, transmembrane transport, chromatin silencing, and some genes harbored unknown functions that all together may respond to salinity as a complex network. For genomic prediction (GP), the genomic best linear unbiased prediction (GBLUP) model reflected genetic effects better than both bayesian ridge regression (BRR) and ridge regression-best linear unbiased prediction (RRBLUP), suggesting GBLUP as a favorable tool for wheat genomic selection.ConclusionThe SNPs and candidate genes identified in the current work can be used potentially for developing salt-tolerant varieties at the seedling growth stage by marker-assisted selection.

Genetic variation and marker-trait association affect the genomic selection prediction accuracy of soybean protein and oil content.

Genomic selection (GS) is a potential breeding approach for soybean improvement. In this study, GS was performed on soybean protein and oil content using the Ridge Regression Best Linear Unbiased Predictor (RR-BLUP) based on 1,007 soybean accessions. The SoySNP50K SNP dataset of the accessions was obtained from the USDA-ARS, Beltsville, MD lab, and the protein and oil content of the accessions were obtained from GRIN. Our results showed that the prediction accuracy of oil content was higher than that of protein content. When the training population size was 100, the prediction accuracies for protein content and oil content were 0.60 and 0.79, respectively. The prediction accuracy increased with the size of the training population. Training populations with similar phenotype or with close genetic relationships to the prediction population exhibited better prediction accuracy. A greatest prediction accuracy for both protein and oil content was observed when approximately 3,000 markers with -log10(P) greater than 1 were included. This information will help improve GS efficiency and facilitate the application of GS.

DETECTING EPISTATIC EFFECTS UNDER A QUALITATIVE CONTEXT: IMPORTANCE AND QUANTIFICATION

The objectives of this work were to simulate and quantify epistatic effects on oligogenic traits and to verify the efficiency of Artificial Neural Networks (ANN) and Ridge Regression Best Linear Unbiased Predictor (RRBLUP) in the prediction of genetic values of oligogenic traits controlled by epistatic genes. We simulated 10 F2 populations in Hardy-Weinberg equilibrium with 800 individuals, each. The individuals were genotyped with 105 codominant markers equidistantly distributed along 10 chromosomes. Genotypic values were simulated for each individual considering five oligogenic traits controlled by two biallelic loci, according to five different epistatic models: duplicate recessive genes, dominant and recessive interaction, duplicate dominant genes, recessive epistasis, and dominant epistasis. RRBLUP, as well as an ANN, were used to perform genomic selection. The coefficient of determination of the regression model revealed a mean epistatic effect ranging from 13.3% in the duplicate recessive genes to 62.5% in the duplicate dominant genes. The identification of epistatic genes was superior in the ANN model compared with the RRBLUP approach. Our result reinforces the potential of ANN in predicting genetic values in situations where other than linear or quadratic relationships are present. The results presented in this work offer important insights about the exploration of epistasis in the qualitative context. In situations where epistatic effects are completely ignored, they can play a role on genetic values, as seen for the duplicative dominant genes.

Evaluation of Genomic Selection Methods for Wheat Quality Traits in Biparental Populations Indicates Inclination towards Parsimonious Solutions

Breeding for end-use quality traits is often challenging since their assessment requires larger quantities of grain and flour samples, which are usually not available early in the breeding process. Using the mixograph as a fast and effective method of evaluating dough quality together with genomic selection (GS) can help in pre-selecting high-performing progenies earlier in the breeding process and achieve a higher gain per unit of time and cost. In the present study, the potential of GS to predict seven end-use quality traits, including mixograph traits, in two biparental wheat populations was investigated. Field trials with both populations were conducted at two locations in Croatia (Osijek and Slavonski Brod) over three years. Results showed that the size of the training population (TP) plays an important role in achieving higher prediction accuracies, while marker density is not a major limitation. Additionally, results of the present study did not support the optimization of TP based on phenotypic variance as a tool to increase prediction accuracy. The performance of eight prediction models was compared and among them elastic net showed the lowest prediction accuracy for all traits. Bayesian models provided slightly higher prediction accuracy than the ridge regression best linear unbiased prediction (RR-BLUP) model, which is negligible considering the time required to perform an analysis. Although RR-BLUP was not the best performing model in all cases, no advantage of using any other model studied here was observed. Furthermore, strong differences between environments in terms of the prediction accuracy achieved were observed, suggesting that environments that are less predictive should be removed from the dataset used to train the prediction model. The prediction accuracies obtained in this study support implementation of GS in wheat breeding for end-use quality, including some mixograph traits.

Response to Early Generation Genomic Selection for Yield in Wheat.

We investigated increasing genetic gain for grain yield using early generation genomic selection (GS). A training set of 1,334 elite wheat breeding lines tested over three field seasons was used to generate Genomic Estimated Breeding Values (GEBVs) for grain yield under irrigated conditions applying markers and three different prediction methods: (1) Genomic Best Linear Unbiased Predictor (GBLUP), (2) GBLUP with the imputation of missing genotypic data by Ridge Regression BLUP (rrGBLUP_imp), and (3) Reproducing Kernel Hilbert Space (RKHS) a.k.a. Gaussian Kernel (GK). F2 GEBVs were generated for 1,924 individuals from 38 biparental cross populations between 21 parents selected from the training set. Results showed that F2 GEBVs from the different methods were not correlated. Experiment 1 consisted of selecting F2s with the highest average GEBVs and advancing them to form genomically selected bulks and make intercross populations aiming to combine favorable alleles for yield. F4:6 lines were derived from genomically selected bulks, intercrosses, and conventional breeding methods with similar numbers from each. Results of field-testing for Experiment 1 did not find any difference in yield with genomic compared to conventional selection. Experiment 2 compared the predictive ability of the different GEBV calculation methods in F2 using a set of single plant-derived F2:4 lines from randomly selected F2 plants. Grain yield results from Experiment 2 showed a significant positive correlation between observed yields of F2:4 lines and predicted yield GEBVs of F2 single plants from GK (the predictive ability of 0.248, P < 0.001) and GBLUP (0.195, P < 0.01) but no correlation with rrGBLUP_imp. Results demonstrate the potential for the application of GS in early generations of wheat breeding and the importance of using the appropriate statistical model for GEBV calculation, which may not be the same as the best model for inbreds.

Deep Learning Algorithms for Complex Traits Genomic Prediction

The underlying perception of genomic selection (GS) is to use genome-wide from DNA sequence (“SNP markers”) along with phenotypes from an observed population to make prediction for the phenotypic outcomes of untested individuals in crop and livestock breeding programs. GS was firstly described by Meuwissen et al.(2001) in dairy cattle to identify genetically superior animals at an early age. The aim was to capture specific genes across the whole genome that are associated with desired traits. The major challenge in using GS programs is to predict the effect of many SNP markers using phenotypic information from a few individuals (aka small n big p problem, or p &amp;gt;&amp;gt; n). Many approaches including naïve and scaled elastic net, ridge regression BLUP Bayesian approaches (BayesA, BayesB, BayesCπ, BayesDπ) LASSO, Support Vector Regression have been conducted to address the small n big p (aka, p &amp;gt;&amp;gt; n) problem. These methods all perform well for (p&amp;gt;&amp;gt;n) by using linear approximation to set a functional relationship between genotypes and phenotypes. However, these methods may not fully capture non-linear effects which are possible to be crucial for complex traits. To deal with this limitation, many methods including neural networks (NN) were recommended to cover non-linearity for GS. Artificial NNs (ANNs) for GS was first presented by Okut et al. (2011) who establish a fully connected regularized multi-layer ANN (MLANN) comprising one hidden layer to predict the body mass index (BMI) in mice using dense molecular markers. Since then, rather complex ANNs approaches have been applied including deep learning (DL) networks. The different DL algorithms have their own advantages to deal with specific problems in complex trait GS. Four different major classes of DL approaches such as fully connected deep learning artificial neural networks (DL-MLANN), recurrent neural networks (RNN), convolutional neural networks (CNN) and long-short term memory (LSTM) and some variation of these network architectures will be summarized here.

Multi-model genome-wide association and genomic prediction analysis of 16 agronomic, physiological and quality related traits in ICARDA spring wheat

Identification and exploration of the genetic architecture of traits related to yield, quality, and drought and heat tolerance is important for yield and quality improvement of wheat through marker-assisted selection. One hundred and ninety-two spring wheat genotypes were tested at two heat-stress locations in Sudan (Wad Medani and Dongula), a drought stress site in Morocco (Marchouch) and a site with high yield potential in Egypt (Sids) in replicated trials during the 2015–2016 and 2016–2017 cropping seasons. A total of 10,577 single nucleotide polymorphism markers identified from the 15 K wheat SNP assay were used in a genome-wide association (GWA) study and genomic prediction for 16 phenotypic traits related to yield, quality and drought and heat tolerance. Significant marker-trait associations were detected across GWAS models for all traits. Most detected marker-trait associations (MTAs) were environment-specific, signifying the presence of high quantitative trait loci-by-environment (QTL x E) interaction. Chromosome arm 5AL had significant multi-model MTAs for grain yield and yield-related traits at the heat-stress locations. Highly significant QTLs were detected on chromosome 2D for waxiness. Homoeologous group 2 and 6 chromosomes were with significant MTAs for grain protein content, gluten content, alveograph strength and Zeleny sedimentation test while chromosome arm 3BL was significant for both Z and W traits. Genomic prediction analysis with ridge regression-best linear unbiased prediction model estimated the breeding values of the studied traits with prediction accuracies ranging from 0.16 for leaf rolling to 0.72 for peduncle length. The identified QTLs could be targeted for marker-assisted selection or further studies aimed at fine mapping and cloning the causative genes and detecting favorable haplotypes with positive effects for agronomic, physiological or quality-related traits.

Evaluations of Genomic Prediction and Identification of New Loci for Resistance to Stripe Rust Disease in Wheat (Triticum aestivum L.).

Stripe rust is one of the most destructive diseases of wheat (Triticum aestivum L.), caused by Puccinia striiformis f. sp. tritici (Pst), and responsible for significant yield losses worldwide. Single-nucleotide polymorphism (SNP) diagnostic markers were used to identify new sources of resistance at adult plant stage to wheat stripe rust (YR) in 141 CIMMYT advanced bread wheat lines over 3 years in replicated trials at Borlaug Institute for South Asia (BISA), Ludhiana. We performed a genome-wide association study and genomic prediction to aid the genetic gain by accumulating disease resistance alleles. The responses to YR in 141 advanced wheat breeding lines at adult plant stage were used to generate G × E (genotype × environment)-dependent rust scores for prediction and genome-wide association study (GWAS), eliminating variation due to climate and disease pressure changes. The lowest mean prediction accuracies were 0.59 for genomic best linear unbiased prediction (GBLUP) and ridge-regression BLUP (RRBLUP), while the highest mean was 0.63 for extended GBLUP (EGBLUP) and random forest (RF), using 14,563 SNPs and the G × E rust score results. RF and EGBLUP predicted higher accuracies (∼3%) than did GBLUP and RRBLUP. Promising genomic prediction demonstrates the viability and efficacy of improving quantitative rust tolerance. The resistance to YR in these lines was attributed to eight quantitative trait loci (QTLs) using the FarmCPU algorithm. Four (Q.Yr.bisa-2A.1, Q.Yr.bisa-2D, Q.Yr.bisa-5B.2, and Q.Yr.bisa-7A) of eight QTLs linked to the diagnostic markers were mapped at unique loci (previously unidentified for Pst resistance) and possibly new loci. The statistical evidence of effectiveness and distribution of the new diagnostic markers for the resistance loci would help to develop new stripe rust resistance sources. These diagnostic markers along with previously established markers would be used to create novel DNA biosensor-based microarrays for rapid detection of the resistance loci on large panels upon functional validation of the candidate genes identified in the present study to aid in rapid genetic gain in the future breeding programs.

Genomic Selection for End-Use Quality and Processing Traits in Soft White Winter Wheat Breeding Program with Machine and Deep Learning Models.

Simple SummaryWheat (Triticum aestivum L.) breeding programs mainly focus on improving grain yield, biotic and abiotic stress tolerance, and end-use quality traits. End-use quality and processing traits are the combinations of various predefined parameters. Genomic selection (GS) opens the potential for selecting improved end-use quality lines. This study explored the potential of the machine and deep learning-based GS models for predicting end-use quality traits in wheat. Deep models were superior to traditional statistical and Bayesian models under all the prediction scenarios. The high accuracy observed for end-use quality traits in this study support predicting them in early generations, leading to the advancement of superior genotypes to more extensive grain yield trails.Breeding for grain yield, biotic and abiotic stress resistance, and end-use quality are important goals of wheat breeding programs. Screening for end-use quality traits is usually secondary to grain yield due to high labor needs, cost of testing, and large seed requirements for phenotyping. Genomic selection provides an alternative to predict performance using genome-wide markers under forward and across location predictions, where a previous year’s dataset can be used to build the models. Due to large datasets in breeding programs, we explored the potential of the machine and deep learning models to predict fourteen end-use quality traits in a winter wheat breeding program. The population used consisted of 666 wheat genotypes screened for five years (2015–19) at two locations (Pullman and Lind, WA, USA). Nine different models, including two machine learning (random forest and support vector machine) and two deep learning models (convolutional neural network and multilayer perceptron) were explored for cross-validation, forward, and across locations predictions. The prediction accuracies for different traits varied from 0.45–0.81, 0.29–0.55, and 0.27–0.50 under cross-validation, forward, and across location predictions. In general, forward prediction accuracies kept increasing over time due to increments in training data size and was more evident for machine and deep learning models. Deep learning models were superior over the traditional ridge regression best linear unbiased prediction (RRBLUP) and Bayesian models under all prediction scenarios. The high accuracy observed for end-use quality traits in this study support predicting them in early generations, leading to the advancement of superior genotypes to more extensive grain yield trails. Furthermore, the superior performance of machine and deep learning models strengthens the idea to include them in large scale breeding programs for predicting complex traits.

A classic approach for determining genomic prediction accuracy under terminal drought stress and well-watered conditions in wheat landraces and cultivars.

The present study aimed to improve the accuracy of genomic prediction of 16 agronomic traits in a diverse bread wheat (Triticum aestivum L.) germplasm under terminal drought stress and well-watered conditions in semi-arid environments. An association panel including 87 bread wheat cultivars and 199 landraces from Iran bread wheat germplasm was planted under two irrigation systems in semi-arid climate zones. The whole association panel was genotyped with 9047 single nucleotide polymorphism markers using the genotyping-by-sequencing method. A number of 23 marker-trait associations were selected for traits under each condition, whereas 17 marker-trait associations were common between terminal drought stress and well-watered conditions. The identified marker-trait associations were mostly single nucleotide polymorphisms with minor allele effects. This study examined the effect of population structure, genomic selection method (ridge regression-best linear unbiased prediction, genomic best-linear unbiased predictions, and Bayesian ridge regression), training set size, and type of marker set on genomic prediction accuracy. The prediction accuracies were low (-0.32) to moderate (0.52). A marker set including 93 significant markers identified through genome-wide association studies with P values ≤ 0.001 increased the genomic prediction accuracy for all traits under both conditions. This study concluded that obtaining the highest genomic prediction accuracy depends on the extent of linkage disequilibrium, the genetic architecture of trait, genetic diversity of the population, and the genomic selection method. The results encouraged the integration of genome-wide association study and genomic selection to enhance genomic prediction accuracy in applied breeding programs.

Artificial Neural Networks in the Prediction of Genetic Merit to Flowering Traits in Bean Cultivars

Flowering is an important agronomic trait that presents non-additive gene action. Genome-enabled prediction allow incorporating molecular information into the prediction of individual genetic merit. Artificial neural networks (ANN) recognize patterns of data and represent an alternative as a universal approximation of complex functions. In a Genomic Selection (GS) context, the ANN allows automatically to capture complicated factors such as epistasis and dominance. The objectives of this study were to predict the individual genetic merits of the traits associated with the flowering time in the common bean using the ANN approach, and to compare the predictive abilities obtained for ANN and Ridge Regression Best Linear Unbiased Predictor (RR-BLUP). We used a set of 80 bean cultivars and genotyping was performed with a set of 384 SNPs. The higher accuracy of the selective process of phenotypic values based on ANN output values resulted in a greater efficacy of the genomic estimated breeding value (GEBV). Through the root mean square error computational intelligence approaches via ANN, GEBV were shown to have greater efficacy than GS via RR-BLUP.

A simple, cost-effective high-throughput image analysis pipeline improves genomic prediction accuracy for days to maturity in wheat

BackgroundHigh-throughput phenotyping and genomic selection accelerate genetic gain in breeding programs by advances in phenotyping and genotyping methods. This study developed a simple, cost-effective high-throughput image analysis pipeline to quantify digital images taken in a panel of 286 Iran bread wheat accessions under terminal drought stress and well-watered conditions. The color proportion of green to yellow (tolerance ratio) and the color proportion of yellow to green (stress ratio) was assessed for each canopy using the pipeline. The estimated tolerance and stress ratios were used as covariates in the genomic prediction models to evaluate the effect of change in canopy color on the improvement of the genomic prediction accuracy of different agronomic traits in wheat.ResultsThe reliability of the high-throughput image analysis pipeline was proved by three to four times of improvement in the accuracy of genomic predictions for days to maturity with the use of tolerance and stress ratios as covariates in the univariate genomic selection models. The higher prediction accuracies were attained for days to maturity when both tolerance and stress ratios were used as fixed effects in the univariate models. The results of this study indicated that the Bayesian ridge regression and ridge regression-best linear unbiased prediction methods were superior to other genomic prediction methods which were used in this study under terminal drought stress and well-watered conditions, respectively.ConclusionsThis study provided a robust, quick, and cost-effective machine learning-enabled image-phenotyping pipeline to improve the genomic prediction accuracy for days to maturity in wheat. The results encouraged the integration of phenomics and genomics in breeding programs.

Natural variation and genomic prediction of growth, physiological traits, and nitrogen-use efficiency in perennial ryegrass under low-nitrogen stress.

Genomic prediction of nitrogen-use efficiency (NUE) has not previously been studied in perennial grass species exposed to low-N stress. Here, we conducted a genomic prediction of physiological traits and NUE in 184 global accessions of perennial ryegrass (Lolium perenne) in response to a normal (7.5 mM) and low (0.75 mM) supply of N. After 21 d of treatment under greenhouse conditions, significant variations in plant height increment (ΔHT), leaf fresh weight (LFW), leaf dry weight (LDW), chlorophyll index (Chl), chlorophyll fluorescence, leaf N and carbon (C) contents, C/N ratio, and NUE were observed in accessions , but to a greater extent under low-N stress. Six genomic prediction models were applied to the data, namely the Bayesian method Bayes C, Bayesian LASSO, Bayesian Ridge Regression, Ridge Regression-Best Linear Unbiased Prediction, Reproducing Kernel Hilbert Spaces, and randomForest. These models produced similar prediction accuracy of traits within the normal or low-N treatments, but the accuracy differed between the two treatments. ΔHT, LFW, LDW, and C were predicted slightly better under normal N with a mean Pearson r-value of 0.26, compared with r=0.22 under low N, while the prediction accuracies for Chl, N, C/N, and NUE were significantly improved under low-N stress with a mean r=0.45, compared with r=0.26 under normal N. The population panel contained three population structures, which generally had no effect on prediction accuracy. The moderate prediction accuracies obtained for N, C, and NUE under low-N stress are promising, and suggest a feasible means by which germplasm might be initially assessed for further detailed studies in breeding programs.

Linkage disequilibrium vs. pedigree: Genomic selection prediction accuracy in conifer species.

BackgroundThe presupposition of genomic selection (GS) is that predictive accuracies should be based on population-wide linkage disequilibrium (LD). However, in species with large, highly complex genomes the limitation of marker density may preclude the ability to resolve LD accurately enough for GS. Here we investigate such an effect in two conifer species with ~ 20 Gbp genomes, Douglas-fir (Pseudotsuga menziesii Mirb. (Franco)) and Interior spruce (Picea glauca (Moench) Voss x Picea engelmannii Parry ex Engelm.). Random sampling of markers was performed to obtain SNP sets with totals in the range of 200–50,000, this was replicated 10 times. Ridge Regression Best Linear Unbiased Predictor (RR-BLUP) was deployed as the GS method to test these SNP sets, and 10-fold cross-validation was performed on 1,321 Douglas-fir trees, representing 37 full-sib F1 families and on 1,126 Interior spruce trees, representing 25 open-pollinated (half-sib) families. Both trials are located on 3 sites in British Columbia, Canada.ResultsAs marker number increased, so did GS predictive accuracy for both conifer species. However, a plateau in the gain of accuracy became apparent around 10,000–15,000 markers for both Douglas-fir and Interior spruce. Despite random marker selection, little variation in predictive accuracy was observed across replications. On average, Douglas-fir prediction accuracies were higher than those of Interior spruce, reflecting the difference between full- and half-sib families for Douglas-fir and Interior spruce populations, respectively, as well as their respective effective population size.ConclusionsAlthough possibly advantageous within an advanced breeding population, reducing marker density cannot be recommended for carrying out GS in conifers. Significant LD between markers and putative causal variants was not detected using 50,000 SNPS, and GS was enabled only through the tracking of relatedness in the populations studied. Dramatically increasing marker density would enable said markers to better track LD with causal variants in these large, genetically diverse genomes; as well as providing a model that could be used across populations, breeding programs, and traits.

Genome-wide association study of milk production traits in a crossbred dairy sheep population using three statistical models.

Milk production is one of the most important characteristics of dairy sheep, and the identification of genes affecting milk production traits is critical to understanding the genetics and improve milk production in future generations. Three statistical techniques, namely GWAS, ridge-regression BLUP and BayesC , were used to identify SNPs in significant association with three milk production traits (milk yield, fat yield and protein yield) in a crossbred dairy sheep population. The results suggested that chromosomes 1, 3, 4, 5, 7 and 11 were likely to harbor genes important to milk production because these chromosomes had the greatest top-100-SNP variance contributions on the three milk production traits. The GWAS analysis identified between 74 and 288 genome-wide significant SNP (P<0.05) whereas the BayesCπ model revealed between six and 63 SNPs, each with >95% posterior probability of inclusion as having a non-zero association effect on at least one of the three milk production traits. Positional candidate genes for milk production in sheep were searched, based on the sheep genomic assembly OAR version 3.1, such as those which map position coincided with or was located within 0.1 Mbp of a genome-wide suggestive or significant SNP. These identified SNPs and candidate genes supported some previous findings and also added new information about genetic markers for genetic improvement of lactation in dairy sheep, but keeping in mind that the majority of these positional candidate genes are not necessarily true causative loci for these traits and future validations are thus necessary.