Genomic evidence for the suitability of Göttingen Minipigs with a rare seizure phenotype as a model for human epilepsy.

Epilepsy is a complex genetic disorder that affects about 2% of the global population. Although the frequency and severity of epileptic seizures can be reduced by a range of pharmacological interventions, there are no disease-modifying treatments for epilepsy. The development of new and more effective drugs is hindered by a lack of suitable animal models. Available rodent models may not recapitulate all key aspects of the disease. Spontaneous epileptic convulsions were observed in few Göttingen Minipigs (GMPs), which may provide a valuable alternative animal model for the characterisation of epilepsy-type diseases and for testing new treatments. We have characterised affected GMPs at the genome level and have taken advantage of primary fibroblast cultures to validate the functional impact of fixed genetic variants on the transcriptome level. We found numerous genes connected to calcium metabolism that have not been associated with epilepsy before, such as ADORA2B, CAMK1D, ITPKB, MCOLN2, MYLK, NFATC3, PDGFD, and PHKB. Our results have identified two transcription factor genes, EGR3 and HOXB6, as potential key regulators of CACNA1H, which was previously linked to epilepsy-type disorders in humans. Our findings provide the first set of conclusive results to support the use of affected subsets of GMPs as an alternative and more reliable model system to study human epilepsy. Further neurological and pharmacological validation of the suitability of GMPs as an epilepsy model is therefore warranted.

Optimization of breeding program design through stochastic simulation with kernel regression.

In recent years, breeding programs have increased significantly in size and complexity, with various highly interdependent parameters and many contrasting breeding goals. As a result, resource allocation in these programs has become more complex, and deriving an optimal breeding strategy has become increasingly challenging. To address this, a common practice is to reduce the optimization problem to a set of scenarios that differ only in a few parameters and can therefore be analyzed in detail. The goal of this article is to provide a framework for the numerical optimization of breeding programs that goes beyond the simple comparison of scenarios. For this, we first determine the space of potential breeding programs only limited by basic constraints like the budget and housing capacities. Subsequently, the goal is to identify the optimal breeding program by finding the parametrization that maximizes the target function by combining different breeding goals. To assess the value of the target function for a parametrization, we propose using stochastic simulations and the subsequent use of a kernel regression method to cope with the stochasticity of simulation outcomes. This procedure is performed iteratively to narrow down the most promising areas of the search space and perform more and more simulations in these areas of interest. In a simplified example applied to a dairy cattle program, our proposed framework has shown its ability to identify an optimal breeding strategy that aligns with a target function aiming at genetic gain and genetic diversity conservation limited by budget constraints.

How economic weights translate into genetic and phenotypic progress, and vice versa.

BACKGROUND: This paper highlights the relationships between economic weights, genetic progress, and phenotypic progress in genomic breeding programs that aim at generating genetic progress in complex, i.e., multi-trait, breeding objectives via a combination of estimated breeding values for different trait complexes. RESULTS: Based on classical selection index theory in combination with quantitative genetic models, we provide a methodological framework for calculating expected genetic and phenotypic progress for all components of a complex breeding objective. We further provide an approach to study the sensitivity of the system to modifications, e.g. to changes in the economic weights. We propose a novel approach to derive the covariance structure of the stochastic errors of estimated breeding values from the observed correlations of estimated breeding values. We define 'realized economic weights' as those weights that would coincide with the observed composition of the genetic trend and show, how they can be calculated. The suggested methodology is illustrated with an index that aims at achieving a breeding goal composed of six trait complexes, that was applied in German Holstein cattle breeding until 2021. CONCLUSIONS: Based on the presented results, the main conclusions are (i) the composition of the observed genetic progress matches the expectations well, with predictions being slightly better when the covariance of estimation errors is taken into account; (ii) the composition of the expected phenotypic trend deviates significantly from the expected genetic trend due to the differences in trait heritabilities; and (iii) the realized economic weights derived from the observed genetic trend deviate substantially from the predefined ones, in one case even with a reversed sign. Further results highlight the implications of the change to a modified breeding goal based on the example of a new index comprising eight, partly new, trait complexes, which is used since 2021 in the German Holstein breeding program. The proposed framework and the analytical tools and software provided will be useful to define more rational and generally accepted breeding objectives in the future.

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.

Ghat: an R package for identifying adaptive polygenic traits.

Identifying selection on polygenic complex traits in crops and livestock is important for understanding evolution and helps prioritize important characteristics for breeding. Quantitative trait loci (QTL) that contribute to polygenic trait variation often exhibit small or infinitesimal effects. This hinders the ability to detect QTL-controlling polygenic traits because enormously high statistical power is needed for their detection. Recently, we circumvented this challenge by introducing a method to identify selection on complex traits by evaluating the relationship between genome-wide changes in allele frequency and estimates of effect size. The approach involves calculating a composite statistic across all markers that capture this relationship, followed by implementing a linkage disequilibrium-aware permutation test to evaluate if the observed pattern differs from that expected due to drift during evolution and population stratification. In this manuscript, we describe "Ghat," an R package developed to implement this method to test for selection on polygenic traits. We demonstrate the package by applying it to test for polygenic selection on 15 published European wheat traits including yield, biomass, quality, morphological characteristics, and disease resistance traits. Moreover, we applied Ghat to different simulated populations with different breeding histories and genetic architectures. The results highlight the power of Ghat to identify selection on complex traits. The Ghat package is accessible on CRAN, the Comprehensive R Archival Network, and on GitHub.

learnMET: an R package to apply machine learning methods for genomic prediction using multi-environment trial data.

We introduce the R-package learnMET, developed as a flexible framework to enable a collection of analyses on multi-environment trial breeding data with machine learning-based models. learnMET allows the combination of genomic information with environmental data such as climate and/or soil characteristics. Notably, the package offers the possibility of incorporating weather data from field weather stations, or to retrieve global meteorological datasets from a NASA database. Daily weather data can be aggregated over specific periods of time based on naive (for instance, nonoverlapping 10-day windows) or phenological approaches. Different machine learning methods for genomic prediction are implemented, including gradient-boosted decision trees, random forests, stacked ensemble models, and multilayer perceptrons. These prediction models can be evaluated via a collection of cross-validation schemes that mimic typical scenarios encountered by plant breeders working with multi-environment trial experimental data in a user-friendly way. The package is published under an MIT license and accessible on GitHub.

Epigenetic Regulation of Phenotypic Sexual Plasticity Inducing Skewed Sex Ratio in Zebrafish.

The plasticity of sexual phenotype in response to environmental conditions results in biased sex ratios, and their variation has an effect on population dynamics. Epigenetic modifications can modulate sex ratio variation in species, where sex is determined by genetic and environmental factors. However, the role of epigenetic mechanisms underlying skewed sex ratios is far from being clear and is still an object of debate in evolutionary developmental biology. In this study, we used zebrafish as a model animal to investigate the effect of DNA methylation on sex ratio variation in sex-biased families in response to environmental temperature. Two sex-biased families with a significant difference in sex ratio were selected for genome-wide DNA methylation analysis using reduced representation bisulfite sequencing (RRBS). The results showed significant genome-wide methylation differences between male-biased and female-biased families, with a greater number of methylated CpG sites in testes than ovaries. Likewise, pronounced differences between testes and ovaries were identified within both families, where the male-biased family exhibited a higher number of methylated sites than the female-biased family. The effect of temperature showed more methylated positions in the high incubation temperature than the control temperature. We found differential methylation of many reproduction-related genes (e.g., sox9a, nr5a2, lhx8a, gata4) and genes involved in epigenetic mechanisms (e.g., dnmt3bb.1, dimt1l, hdac11, h1m) in both families. We conclude that epigenetic modifications can influence the sex ratio variation in zebrafish families and may generate skewed sex ratios, which could have a negative consequence for population fitness in species with genotype-environment interaction sex-determining system under rapid environmental changes.

Assessment of linkage disequilibrium patterns between structural variants and single nucleotide polymorphisms in three commercial chicken populations.

BACKGROUND: Structural variants (SV) are causative for some prominent phenotypic traits of livestock as different comb types in chickens or color patterns in pigs. Their effects on production traits are also increasingly studied. Nevertheless, accurately calling SV remains challenging. It is therefore of interest, whether close-by single nucleotide polymorphisms (SNPs) are in strong linkage disequilibrium (LD) with SVs and can serve as markers. Literature comes to different conclusions on whether SVs are in LD to SNPs on the same level as SNPs to other SNPs. The present study aimed to generate a precise SV callset from whole-genome short-read sequencing (WGS) data for three commercial chicken populations and to evaluate LD patterns between the called SVs and surrounding SNPs. It is thereby the first study that assessed LD between SVs and SNPs in chickens. RESULTS: The final callset consisted of 12,294,329 bivariate SNPs, 4,301 deletions (DEL), 224 duplications (DUP), 218 inversions (INV) and 117 translocation breakpoints (BND). While average LD between DELs and SNPs was at the same level as between SNPs and SNPs, LD between other SVs and SNPs was strongly reduced (DUP: 40%, INV: 27%, BND: 19% of between-SNP LD). A main factor for the reduced LD was the presence of local minor allele frequency differences, which accounted for 50% of the difference between SNP - SNP and DUP - SNP LD. This was potentially accompanied by lower genotyping accuracies for DUP, INV and BND compared with SNPs and DELs. An evaluation of the presence of tag SNPs (SNP in highest LD to the variant of interest) further revealed DELs to be slightly less tagged by WGS SNPs than WGS SNPs by other SNPs. This difference, however, was no longer present when reducing the pool of potential tag SNPs to SNPs located on four different chicken genotyping arrays. CONCLUSIONS: The results implied that genomic variance due to DELs in the chicken populations studied can be captured by different SNP marker sets as good as variance from WGS SNPs, whereas separate SV calling might be advisable for DUP, INV, and BND effects.

Prediction of Maize Phenotypic Traits With Genomic and Environmental Predictors Using Gradient Boosting Frameworks.

The development of crop varieties with stable performance in future environmental conditions represents a critical challenge in the context of climate change. Environmental data collected at the field level, such as soil and climatic information, can be relevant to improve predictive ability in genomic prediction models by describing more precisely genotype-by-environment interactions, which represent a key component of the phenotypic response for complex crop agronomic traits. Modern predictive modeling approaches can efficiently handle various data types and are able to capture complex nonlinear relationships in large datasets. In particular, machine learning techniques have gained substantial interest in recent years. Here we examined the predictive ability of machine learning-based models for two phenotypic traits in maize using data collected by the Maize Genomes to Fields (G2F) Initiative. The data we analyzed consisted of multi-environment trials (METs) dispersed across the United States and Canada from 2014 to 2017. An assortment of soil- and weather-related variables was derived and used in prediction models alongside genotypic data. Linear random effects models were compared to a linear regularized regression method (elastic net) and to two nonlinear gradient boosting methods based on decision tree algorithms (XGBoost, LightGBM). These models were evaluated under four prediction problems: (1) tested and new genotypes in a new year; (2) only unobserved genotypes in a new year; (3) tested and new genotypes in a new site; (4) only unobserved genotypes in a new site. Accuracy in forecasting grain yield performance of new genotypes in a new year was improved by up to 20% over the baseline model by including environmental predictors with gradient boosting methods. For plant height, an enhancement of predictive ability could neither be observed by using machine learning-based methods nor by using detailed environmental information. An investigation of key environmental factors using gradient boosting frameworks also revealed that temperature at flowering stage, frequency and amount of water received during the vegetative and grain filling stage, and soil organic matter content appeared as important predictors for grain yield in our panel of environments.

Increasing calling accuracy, coverage, and read-depth in sequence data by the use of haplotype blocks.

High-throughput genotyping of large numbers of lines remains a key challenge in plant genetics, requiring geneticists and breeders to find a balance between data quality and the number of genotyped lines under a variety of different existing genotyping technologies when resources are limited. In this work, we are proposing a new imputation pipeline ("HBimpute") that can be used to generate high-quality genomic data from low read-depth whole-genome-sequence data. The key idea of the pipeline is the use of haplotype blocks from the software HaploBlocker to identify locally similar lines and subsequently use the reads of all locally similar lines in the variant calling for a specific line. The effectiveness of the pipeline is showcased on a dataset of 321 doubled haploid lines of a European maize landrace, which were sequenced at 0.5X read-depth. The overall imputing error rates are cut in half compared to state-of-the-art software like BEAGLE and STITCH, while the average read-depth is increased to 83X, thus enabling the calling of copy number variation. The usefulness of the obtained imputed data panel is further evaluated by comparing the performance of sequence data in common breeding applications to that of genomic data generated with a genotyping array. For both genome-wide association studies and genomic prediction, results are on par or even slightly better than results obtained with high-density array data (600k). In particular for genomic prediction, we observe slightly higher data quality for the sequence data compared to the 600k array in the form of higher prediction accuracies. This occurred specifically when reducing the data panel to the set of overlapping markers between sequence and array, indicating that sequencing data can benefit from the same marker ascertainment as used in the array process to increase the quality and usability of genomic data.

Genotypic and Dietary Effects on Egg Quality of Local Chicken Breeds and Their Crosses Fed with Faba Beans.

The quality of chicken eggs is an important criterion for food safety and the consumers' choice at the point of sale. Several studies have shown that egg quality can be influenced by the chickens' genotype and by the composition of the diet. The present study aimed to evaluate the effect of faba beans as a substitute for soybeans in the diet of chickens originating from traditional low-performance breeds in comparison with high-performing laying type hens and their crosses on egg quality parameters. Chickens of six different genotypes were fed either with a feed mix containing 20% faba beans with high or low vicin contents or, as a control, a feed mix containing soybeans. The genotypes studied were the local breeds Vorwerkhuhn and Bresse Gauloise, as well as commercial White Rock parent hens and their crosses. Yolk weight, Haugh units, yolk and shell color, the frequency of blood and meat spots and the composition of the eggs were significantly influenced by the genotype. The feeding of faba beans had an effect on yolk and shell color, Haugh units and shell portion, while there was no significant influence on the frequency of blood and meat spots.

In Silico Prediction of Transcription Factor Collaborations Underlying Phenotypic Sexual Dimorphism in Zebrafish (Danio rerio).

The transcriptional regulation of gene expression in higher organisms is essential for different cellular and biological processes. These processes are controlled by transcription factors and their combinatorial interplay, which are crucial for complex genetic programs and transcriptional machinery. The regulation of sex-biased gene expression plays a major role in phenotypic sexual dimorphism in many species, causing dimorphic gene expression patterns between two different sexes. The role of transcription factor (TF) in gene regulatory mechanisms so far has not been studied for sex determination and sex-associated colour patterning in zebrafish with respect to phenotypic sexual dimorphism. To address this open biological issue, we applied bioinformatics approaches for identifying the predicted TF pairs based on their binding sites for sex and colour genes in zebrafish. In this study, we identified 25 (e.g., STAT6-GATA4; JUN-GATA4; SOX9-JUN) and 14 (e.g., IRF-STAT6; SOX9-JUN; STAT6-GATA4) potentially cooperating TFs based on their binding patterns in promoter regions for sex determination and colour pattern genes in zebrafish, respectively. The comparison between identified TFs for sex and colour genes revealed several predicted TF pairs (e.g., STAT6-GATA4; JUN-SOX9) are common for both phenotypes, which may play a pivotal role in phenotypic sexual dimorphism in zebrafish.

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.

How imputation can mitigate SNP ascertainment Bias.

BACKGROUND: Population genetic studies based on genotyped single nucleotide polymorphisms (SNPs) are influenced by a non-random selection of the SNPs included in the used genotyping arrays. The resulting bias in the estimation of allele frequency spectra and population genetics parameters like heterozygosity and genetic distances relative to whole genome sequencing (WGS) data is known as SNP ascertainment bias. Full correction for this bias requires detailed knowledge of the array design process, which is often not available in practice. This study suggests an alternative approach to mitigate ascertainment bias of a large set of genotyped individuals by using information of a small set of sequenced individuals via imputation without the need for prior knowledge on the array design. RESULTS: The strategy was first tested by simulating additional ascertainment bias with a set of 1566 chickens from 74 populations that were genotyped for the positions of the Affymetrix Axiom™ 580 k Genome-Wide Chicken Array. Imputation accuracy was shown to be consistently higher for populations used for SNP discovery during the simulated array design process. Reference sets of at least one individual per population in the study set led to a strong correction of ascertainment bias for estimates of expected and observed heterozygosity, Wright's Fixation Index and Nei's Standard Genetic Distance. In contrast, unbalanced reference sets (overrepresentation of populations compared to the study set) introduced a new bias towards the reference populations. Finally, the array genotypes were imputed to WGS by utilization of reference sets of 74 individuals (one per population) to 98 individuals (additional commercial chickens) and compared with a mixture of individually and pooled sequenced populations. The imputation reduced the slope between heterozygosity estimates of array data and WGS data from 1.94 to 1.26 when using the smaller balanced reference panel and to 1.44 when using the larger but unbalanced reference panel. This generally supported the results from simulation but was less favorable, advocating for a larger reference panel when imputing to WGS. CONCLUSIONS: The results highlight the potential of using imputation for mitigation of SNP ascertainment bias but also underline the need for unbiased reference sets.

Genetic diversity in global chicken breeds in relation to their genetic distances to wild populations.

BACKGROUND: Migration of a population from its founder population is expected to cause a reduction of its genetic diversity and facilitates differentiation between the population and its founder population, as predicted by the theory of genetic isolation by distance. Consistent with that theory, a model of expansion from a single founder predicts that patterns of genetic diversity in populations can be explained well by their geographic expansion from their founders, which is correlated with genetic differentiation. METHODS: To investigate this in chicken, we estimated the relationship between the genetic diversity of 160 domesticated chicken populations and their genetic distances to wild chicken populations. RESULTS: Our results show a strong inverse relationship, i.e. 88.6% of the variation in the overall genetic diversity of domesticated chicken populations was explained by their genetic distance to the wild populations. We also investigated whether the patterns of genetic diversity of different types of single nucleotide polymorphisms (SNPs) and genes are similar to that of the overall genome. Among the SNP classes, the non-synonymous SNPs deviated most from the overall genome. However, genetic distance to the wild chicken still explained more variation in domesticated chicken diversity across all SNP classes, which ranged from 83.0 to 89.3%. CONCLUSIONS: Genetic distance between domesticated chicken populations and their wild relatives can predict the genetic diversity of the domesticated populations. On the one hand, genes with little genetic variation across populations, regardless of the genetic distance to the wild population, are associated with major functions such as brain development. Changes in such genes may be detrimental to the species. On the other hand, genetic diversity seems to change at a faster rate within genes that are associated with e.g. protein transport and protein and lipid metabolic processes. In general, such genes may be flexible to changes according to the populations' needs. These results contribute to the knowledge of the evolutionary patterns of different functional genomic regions in the chicken.

How array design creates SNP ascertainment bias.

Single nucleotide polymorphisms (SNPs), genotyped with arrays, have become a widely used marker type in population genetic analyses over the last 10 years. However, compared to whole genome re-sequencing data, arrays are known to lack a substantial proportion of globally rare variants and tend to be biased towards variants present in populations involved in the development process of the respective array. This affects population genetic estimators and is known as SNP ascertainment bias. We investigated factors contributing to ascertainment bias in array development by redesigning the Axiom™ Genome-Wide Chicken Array in silico and evaluating changes in allele frequency spectra and heterozygosity estimates in a stepwise manner. A sequential reduction of rare alleles during the development process was shown. This was mainly caused by the identification of SNPs in a limited set of populations and a within-population selection of common SNPs when aiming for equidistant spacing. These effects were shown to be less severe with a larger discovery panel. Additionally, a generally massive overestimation of expected heterozygosity for the ascertained SNP sets was shown. This overestimation was 24% higher for populations involved in the discovery process than not involved populations in case of the original array. The same was observed after the SNP discovery step in the redesign. However, an unequal contribution of populations during the SNP selection can mask this effect but also adds uncertainty. Finally, we make suggestions for the design of specialized arrays for large scale projects where whole genome re-sequencing techniques are still too expensive.

MoBPSweb: A web-based framework to simulate and compare breeding programs.

In this study, we introduce a new web-based simulation framework ("MoBPSweb") that combines a unified language to describe breeding programs with the simulation software MoBPS, standing for "Modular Breeding Program Simulator." Thereby, MoBPSweb provides a flexible environment to log, simulate, evaluate, and compare breeding programs. Inputs can be provided via modules ranging from a Vis.js-based environment for "drawing" the breeding program to a variety of modules to provide phenotype information, economic parameters, and other relevant information. Similarly, results of the simulation study can be extracted and compared to other scenarios via output modules (e.g., observed phenotypes, the accuracy of breeding value estimation, inbreeding rates), while all simulations and downstream analysis are executed in the highly efficient R-package MoBPS.

Phenotype Prediction Under Epistasis.

Reliable methods of phenotype prediction from genomic data play an increasingly important role in many areas of plant and animal breeding. Thus, developing methods that enhance prediction accuracy is of major interest. Here, we provide three methods for this purpose: (1) Genomic Best Linear Unbiased Prediction (GBLUP) as a model just accounting for additive SNP effects; (2) Epistatic Random Regression BLUP (ERRBLUP) as a full epistatic model which incorporates all pairwise SNP interactions, and (3) selective Epistatic Random Regression BLUP (sERRBLUP) as an epistatic model which incorporates a subset of pairwise SNP interactions selected based on their absolute effect sizes or the effect variances, which is computed based on solutions from the ERRBLUP model. We compared the predictive ability obtained from GBLUP, ERRBLUP, and sERRBLUP with genotypes from a publicly available wheat dataset and respective simulated phenotypes. Results showed that sERRBLUP provides a substantial increase in prediction accuracy compared to the other methods when the optimal proportion of SNP interactions is kept in the model, especially when an optimal proportion of SNP interactions is selected based on the SNP interaction effect sizes. All methods described here are implemented in the R-package EpiGP, which is able to process large-scale genomic data in a computationally efficient way.