Accuracy Of Genomic Prediction Research Articles

Genome-wide association mapping to identify genetic loci governing agronomic traits and genomic prediction prospects in tetraploid potatoes

Genome-wide association studies (GWAS) and Genomic Prediction (GP) have become powerful tools for examining genetically complex traits in crops. The present study was undertaken on a diverse potato panel of 222 tetraploid potato cultivars and breeding lines, genotyped using genotyping with a sequencing approach to identify 120,622 high-quality single nucleotide polymorphism (SNP) markers. The phenotype data was observed on 12 agronomic traits, including morphological (qualitative) and tuber yield traits (quantitative). The population structure showed three different sub-populations with varying levels of admixtures in them. The LD decay was observed at ∼5Mb distance and GWAS using the Q+K model identified 46 marker-trait associations (MTAs), including 20 associations for tuber morphological traits and rest 26 for tuber yield traits. The average phenotypic variance captured by the associated SNPs was around 10 % for most traits. The putative candidate genes identified for different traits were cytochrome P450 71A4 for flesh color, phytochrome-B and ATP synthase subunit for tuber number, and receptor-like kinases and MYB transcription factor for tuber yield. We did not get any marker trait association for tuber skin color and chip score. The GP accuracy was moderate (0.3–0.5) to high (>0.5) for most of the traits except crisp score, tuber shape and marketable tubers, which recorded low GP accuracy. The reduced marker density did not show any decline in the GP accuracy of different traits. Overall, the study presents the genetic analysis of a diverse panel with information on genetic control and genomic predictions of complex traits for use in genomics-assisted breeding and genomic selection in potatoes.

Recurrent selection for Fusarium head blight resistance in a durum wheat population

AbstractFusarium head blight (FHB) is a devastating fungal disease of wheat. Since early 1990s, frequent FHB epidemics in major wheat‐growing regions have caused massive economic losses. Developing FHB‐resistant varieties is key to minimize the loss caused by the disease. In durum wheat (Triticum turgidum L. ssp. durum), however, it is challenged by lack of resistant sources. Introgression of resistance genes from wild relatives and bread wheat has resulted in some durum wheat lines with moderate resistance. Due to its complex genetic nature, integrating more resistance genes with moderate and even minor effects promises to provide high and durable FHB resistance. In this study, a base population was developed using diverse resistant lines and elite durum wheat breeding lines as founders and went through three cycles of phenotypic selection from 2019 to 2022, with FHB severity decreased by 34%. Six S0:1 lines in Cycle 2 and Cycle 3 populations exhibited significantly lower FHB severity than the moderately susceptible check variety ND Riveland. The results suggested that recurrent phenotypic selection could effectively improve FHB resistance. Analysis of the founders and Cycle 2 and Cycle 3 populations with 2706 single nucleotide polymorphism markers indicated that there was little decrease in genetic variation due to selection. Genome‐wide association analysis found no markers significantly associated with FHB severity. Genomic prediction accuracies based on cross‐validation for FHB severity, plant height, and days‐to‐flowering were 0.51, 0.69, and 0.61, respectively. Recurrent genomic selection can shorten the select cycle to 4 months and therefore is promising to accelerate genetic improvement of FHB severity.

Genomic predictions for resistance to Aeromonas hydrophila in pacu (Piaractus mesopotamicus)

Infections caused by the bacterium Aeromonas hydrophila result in high mortality outbreaks in pacu (Piaractus mesopotamicus) production. An alternative to increase the resistance to A. hydrophila infection is the implementation of genomic selection programs, which can contribute to the sustainable development of the Brazilian aquaculture. In addition, low-density SNP panels, optimized by genotype imputation, represent a low-cost tool for implementing genomic selection on a commercial scale. This study aimed to evaluate the accuracy of genomic prediction using low (0.5 and 1 K) and medium (15 K imputed) density SNP panels; and compare breeding values accuracy with restriction-site associated DNA sequencing (18 K) and pedigree-based methods for individuals challenged for A. hydrophila. The genomic prediction analyses were conducted using genotypes from the medium-density SNP panel of pacu as reference and genotypes from a low-density SNP panel (0.5 and 1 K) using target genotype-by-sequencing technology (tGBS) for imputation. Approximately 15 K SNPs were successfully imputed with accuracy (r2) above 0.8. Time to death (TD) and binary survival (BS) against A. hydrophila infection were used as phenotypes for genomic prediction. Genomic values were estimated using the GBLUP model adopting different genotyping methods (SNP array and RAD-seq) and densities of SNPs (0.5 K, 1 K, and 15K_imputed for the array and 18 K for RAD-seq). Accuracies of genomic approaches using different methods and densities were compared to the pedigree method PBLUP to assess the benefits of incorporating genomic information. The accuracy of predicting breeding values using genomic methods outperformed the estimates obtained through PBLUP by 33%, 39%, and 67% for TD, and 36%, 39%, and 82% for BS, considering the densities of 0.5, 1, and 15K_imputed, respectively. The imputation strategy yielded better results in terms of genomic prediction accuracy for both traits. The RAD-seq genotyping method produced low-accuracy estimates (∼30% lower for both traits) compared to PBLUP, indicating that such method may not be suitable for genomic prediction of these traits. Genomic selection using genotype imputation is a viable and cost-effective approach to increase resistance to A. hydrophila infection in pacu and reduce genotyping costs without compromising genomic prediction accuracy.

Leveraging transcriptomics-based approaches to enhance genomic prediction: integrating SNPs and gene networks for cotton fibre quality improvement.

Cultivated cotton plants are the world's largest source of natural fibre, where yield and quality are key traits for this renewable and biodegradable commodity. The Gossypium hirsutum cotton genome contains ~80K protein-coding genes, making precision breeding of complex traits a challenge. This study tested approaches to improving the genomic prediction (GP) accuracy of valuable cotton fibre traits to help accelerate precision breeding. With a biology-informed basis, a novel approach was tested for improving GP for key cotton fibre traits with transcriptomics of key time points during fibre development, namely, fibre cells undergoing primary, transition, and secondary wall development. Three test approaches included weighting of SNPs in DE genes overall, in target DE gene lists informed by gene annotation, and in a novel approach of gene co-expression network (GCN) clusters created with partial correlation and information theory (PCIT) as the prior information in GP models. The GCN clusters were nucleated with known genes for fibre biomechanics, i.e., fasciclin-like arabinogalactan proteins, and cluster size effects were evaluated. The most promising improvements in GP accuracy were achieved by using GCN clusters for cotton fibre elongation by 4.6%, and strength by 4.7%, where cluster sizes of two and three neighbours proved most effective. Furthermore, the improvements in GP were due to only a small number of SNPs, in the order of 30 per trait using the GCN cluster approach. Non-trait-specific biological time points, and genes, were found to have neutral effects, or even reduced GP accuracy for certain traits. As the GCN clusters were generated based on known genes for fibre biomechanics, additional candidate genes were identified for fibre elongation and strength. These results demonstrate that GCN clusters make a specific and unique contribution in improving the GP of cotton fibre traits. The findings also indicate that there is room for incorporating biology-based GCNs into GP models of genomic selection pipelines for cotton breeding to help improve precision breeding of target traits. The PCIT-GCN cluster approach may also hold potential application in other crops and trees for enhancing breeding of complex traits.

Use of whole-genome sequence data for fine mapping and genomic prediction of sea louse resistance in Atlantic salmon.

Sea louse (Lepeophtheirus salmonis) infestation of Atlantic salmon (Salmo salar) is a significant challenge in aquaculture. Over the years, this parasite has developed immunity to medicinal control compounds, and non-medicinal control methods have been proven to be stressful, hence the need to study the genomic architecture of salmon resistance to sea lice. Thus, this research used whole-genome sequence (WGS) data to study the genetic basis of the trait since most research using fewer SNPs did not identify significant quantitative trait loci. Mowi Genetics AS provided the genotype (50k SNPs) and phenotype data for this research after conducting a sea lice challenge test on 3,185 salmon smolts belonging to 191 full-sib families. The 50 k SNP genotype was imputed to WGS using the information from 197 closely related individuals with sequence data. The WGS and 50k SNPs of the challenged population were then used to estimate genetic parameters, perform a genome-wide association study (GWAS), predict genomic breeding values, and estimate its accuracy for host resistance to sea lice. The heritability of host resistance to sea lice was estimated to be 0.21 and 0.22, while the accuracy of genomic prediction was estimated to be 0.65 and 0.64 for array and WGS data, respectively. In addition, the association test using both array and WGS data did not identify any marker associated with sea lice resistance at the genome-wide level. We conclude that sea lice resistance is a polygenic trait that is moderately heritable. The genomic predictions using medium-density SNP genotyping array were equally good or better than those based on WGS data.

High-Throughput Phenotyping of Soybean Biomass: Conventional Trait Estimation and Novel Latent Feature Extraction Using UAV Remote Sensing and Deep Learning Models.

High-throughput phenotyping serves as a framework to reduce chronological costs and accelerate breeding cycles. In this study, we developed models to estimate the phenotypes of biomass-related traits in soybean (Glycine max) using unmanned aerial vehicle (UAV) remote sensing and deep learning models. In 2018, a field experiment was conducted using 198 soybean germplasm accessions with known whole-genome sequences under 2 irrigation conditions: drought and control. We used a convolutional neural network (CNN) as a model to estimate the phenotypic values of 5 conventional biomass-related traits: dry weight, main stem length, numbers of nodes and branches, and plant height. We utilized manually measured phenotypes of conventional traits along with RGB images and digital surface models from UAV remote sensing to train our CNN models. The accuracy of the developed models was assessed through 10-fold cross-validation, which demonstrated their ability to accurately estimate the phenotypes of all conventional traits simultaneously. Deep learning enabled us to extract features that exhibited strong correlations with the output (i.e., phenotypes of the target traits) and accurately estimate the values of the features from the input data. We considered the extracted low-dimensional features as phenotypes in the latent space and attempted to annotate them based on the phenotypes of conventional traits. Furthermore, we validated whether these low-dimensional latent features were genetically controlled by assessing the accuracy of genomic predictions. The results revealed the potential utility of these low-dimensional latent features in actual breeding scenarios.

Effect of population structure on the accuracy of genomic prediction for carcass traits in Hanwoo cattle: a simulation study

Improvement of quantitative traits in livestock is an essential goal of animal breeding. For the achievement of this goal, an accurate breeding value estimation is needed. The genetic relationship between reference and test groups is a key factor in determining the accuracy of genomic estimated breeding value (GEBV) thus, the structure of the reference population is crucial for efficient genomic selection. By the number of sharing parents, the population structure can be divided into half-sibling and full-sibling families. Also, the population structure of Hanwoo cattle primarily consists of half-sibling families because of the production system. Therefore, comparing half-sibling and full-sibling families is challenging in the Hanwoo cattle population, yet an important issue in the direction of breeding strategy in Korea. The objective of this study was to compare the accuracy of GEBV between different family structures and investigate efficient family size in the reference population using simulated data. 6 different populations were simulated using QMSim software, and the individuals in the last generations were separated into reference and test groups. The GEBV was calculated using BLUPF90 software. Practical accuracy was between 0.36-0.52 in half-sibling families and 0.55-0.77 in full-sibling families. The increase rate of accuracy was highest at the sibling size of 20, with practical accuracy of 0.52 in half-sibling families and 0.77 in full-sibling families. As a result, the most efficient population structure for genomic prediction was a sibling size of 20 in a full-sibling family.

High-throughput field phenotyping reveals that selection in breeding has affected the phenology and temperature response of wheat in the stem elongation phase.

Crop growth and phenology are driven by seasonal changes in environmental variables, with temperature as one important factor. However, knowledge about genotype-specific temperature response and its influence on phenology is limited. Such information is fundamental to improve crop models and adapt selection strategies. We measured the increase in height of 352 European winter wheat varieties in 4 years to quantify phenology, and fitted an asymptotic temperature response model. The model used hourly fluctuations in temperature to parameterize the base temperature (Tmin), the temperature optimum (rmax), and the steepness (lrc) of growth responses. Our results show that higher Tmin and lrc relate to an earlier start and end of stem elongation. A higher rmax relates to an increased final height. Both final height and rmax decreased for varieties originating from the continental east of Europe towards the maritime west. A genome-wide association study (GWAS) indicated a quantitative inheritance and a large degree of independence among loci. Nevertheless, genomic prediction accuracies (GBLUPs) for Tmin and lrc were low (r≤0.32) compared with other traits (r≥0.59). As well as known, major genes related to vernalization, photoperiod, or dwarfing, the GWAS indicated additional, as yet unknown loci that dominate the temperature response.

Genomic Predictions in Korean Hanwoo Cows: A Comparative Analysis of Genomic BLUP and Bayesian Methods for Reproductive Traits.

This study aimed to predict the accuracy of genomic estimated breeding values (GEBVs) for reproductive traits in Hanwoo cows using the GBLUP, BayesB, BayesLASSO, and BayesR methods. Accuracy estimates of GEBVs for reproductive traits were derived through fivefold cross-validation, analyzing a dataset comprising 11,348 animals and employing an Illumina Bovine 50K SNP chip. GBLUP showed an accuracy of 0.26 for AFC, while BayesB, BayesLASSO, and BayesR demonstrated values of 0.28, 0.29, and 0.29, respectively. For CI, GBLUP attained an accuracy of 0.19, whereas BayesB, BayesLASSO, and BayesR scored 0.21, 0.24, and 0.25, respectively. The accuracy for GL was uniform across GBLUP, BayesB, and BayesR at 0.31, whereas BayesLASSO showed a slightly higher accuracy of 0.33. For NAIPC, GBLUP showed an accuracy of 0.24, while BayesB, BayesLASSO, and BayesR recorded 0.22, 0.27, and 0.30, respectively. The variation in genomic prediction accuracy among methods indicated Bayesian approaches slightly outperformed GBLUP. The findings suggest that Bayesian methods, notably BayesLASSO and BayesR, offer improved predictive capabilities for reproductive traits. Future research may explore more advanced genomic approaches to enhance predictive accuracy and genetic gains in Hanwoo cattle breeding programs.

Using genomic selection to improve the accuracy of genomic prediction for multi-populations in pigs

The size of the reference group is among the most critical determinants of genomic estimated breeding values (GEBVs) accuracy. However, small- and medium-sized pig farms often need help accumulating adequate reference data, posing significant challenges to breeding programs. To solve this problem, exploring the potential benefits of combining reference groups of different sizes is necessary to improve GEBV accuracy. The primary objective of this investigation was to assess a more effective statistical model for combined multi-populations and its potential to enhance the accuracy of GEBVs for small and medium populations. Three populations were simulated using the QMSim software, each consisting of different sizes (300, 600, and 1 500, respectively). To assess the impact of heritability on the accuracy of GEBVs, four different levels of heritability (0.05, 0.15, 0.35, and 0.5) were simulated. Simultaneously, to investigate the impact of kinship on multi-populations, the study created four distinct scenarios for the three sizes of populations. These scenarios included: (1) the three groups are all independent, (2) the large group and the small group with a familial connection (n = 1 800), a middle group (n = 600) acting independently with no kinship, (3) the large group with a familial connection to the middle group (n = 2 100) but no connection to the small group (n = 300), and (4) the small group with a familial connection to the middle group (n = 900), while the large group (n = 1 500) acted independently with no kinship. This study evaluates and compares the accuracy of predicting breeding values using four different methods, including genomic best linear unbiased prediction (GBLUP), single-stepGBLUP (ssGBLUP), and two Bayesian models (Bayes A and Bayes B), with varying sizes of reference groups. In each scenario, three different prediction strategies were compared: (1) Merging all three different sizes of populations for predicting, (2) predicting each independent population separately, and (3) the other two populations predict the population. Our findings reveal that combining populations enhances the Bayesian models, with Bayes B yielding the highest accuracy. In independent populations, the best linear unbiased prediction (BLUP) models demonstrated the highest accuracy. However, in cases where populations were related and the heritability was high, the Bayes B model exhibited the highest overall accuracy (slightly higher than BLUP models) in the independent population. Our results underscore the importance of considering population combinations when using genetic models to predict breeding values, particularly for pig farmers with limited resources.

Increased Accuracy of Genomic Prediction Using Preselected SNPs from GWAS with Imputed Whole-Genome Sequence Data in Pigs.

Enhancing the accuracy of genomic prediction is a key goal in genomic selection (GS) research. Integrating prior biological information into GS methods using appropriate models can improve prediction accuracy for complex traits. Genome-wide association study (GWAS) is widely utilized to identify potential candidate loci associated with complex traits in livestock and poultry, offering essential genomic insights. In this study, a GWAS was conducted on 685 Duroc × Landrace × Yorkshire (DLY) pigs to extract significant single-nucleotide polymorphisms (SNPs) as genomic features. We compared two GS models, genomic best linear unbiased prediction (GBLUP) and genomic feature BLUP (GFBLUP), by using imputed whole-genome sequencing (WGS) data on 651 Yorkshire pigs. The results revealed that the GBLUP model achieved prediction accuracies of 0.499 for backfat thickness (BFT) and 0.423 for loin muscle area (LMA). By applying the GFBLUP model with GWAS-based SNP preselection, the average prediction accuracies for BFT and LMA traits reached 0.491 and 0.440, respectively. Specifically, the GFBLUP model displayed a 4.8% enhancement in predicting LMA compared to the GBLUP model. These findings suggest that, in certain scenarios, the GFBLUP model may offer superior genomic prediction accuracy when compared to the GBLUP model, underscoring the potential value of incorporating genomic features to refine GS models.

Genomic loci involved in sensing environmental cues and metabolism affect seasonal coat shedding in Bos taurus and Bos indicus cattle.

Seasonal shedding of winter hair at the start of summer is well studied in wild and domesticated populations. However, the genetic influences on this trait and their interactions are poorly understood. We use data from 13,364 cattle with 36,899 repeated phenotypes to investigate the relationship between hair shedding and environmental variables, single nucleotide polymorphisms, and their interactions to understand quantitative differences in seasonal shedding. Using deregressed estimated breeding values from a repeated records model in a genome-wide association analysis (GWAA) and meta-analysis of year-specific GWAA gave remarkably similar results. These GWAA identified hundreds of variants associated with seasonal hair shedding. There were especially strong associations between chromosomes 5 and 23. Genotype-by-environment interaction GWAA identified 1,040 day length-by-genotype interaction associations and 17 apparent temperature-by-genotype interaction associations with hair shedding, highlighting the importance of day length on hair shedding. Accurate genomic predictions of hair shedding were created for the entire dataset, Angus, Hereford, Brangus, and multibreed datasets. Loci related to metabolism and light-sensing have a large influence on seasonal hair shedding. This is one of the largest genetic analyses of a phenological trait and provides insight into both agriculture production and basic science.

Imputation strategies for genomic prediction using nanopore sequencing

BackgroundGenomic prediction describes the use of SNP genotypes to predict complex traits and has been widely applied in humans and agricultural species. Genotyping-by-sequencing, a method which uses low-coverage sequence data paired with genotype imputation, is becoming an increasingly popular SNP genotyping method for genomic prediction. The development of Oxford Nanopore Technologies’ (ONT) MinION sequencer has now made genotyping-by-sequencing portable and rapid. Here we evaluate the speed and accuracy of genomic predictions using low-coverage ONT sequence data in a population of cattle using four imputation approaches. We also investigate the effect of SNP reference panel size on imputation performance.ResultsSNP array genotypes and ONT sequence data for 62 beef heifers were used to calculate genomic estimated breeding values (GEBVs) from 641 k SNP for four traits. GEBV accuracy was much higher when genome-wide flanking SNP from sequence data were used to help impute the 641 k panel used for genomic predictions. Using the imputation package QUILT, correlations between ONT and low-density SNP array genomic breeding values were greater than 0.91 and up to 0.97 for sequencing coverages as low as 0.1 × using a reference panel of 48 million SNP. Imputation time was significantly reduced by decreasing the number of flanking sequence SNP used in imputation for all methods. When compared to high-density SNP arrays, genotyping accuracy and genomic breeding value correlations at 0.5 × coverage were also found to be higher than those imputed from low-density arrays.ConclusionsHere we demonstrated accurate genomic prediction is possible with ONT sequence data from sequencing coverages as low as 0.1 × , and imputation time can be as short as 10 min per sample. We also demonstrate that in this population, genotyping-by-sequencing at 0.1 × coverage can be more accurate than imputation from low-density SNP arrays.

Preselecting Variants from Large-Scale Genome-Wide Association Study Meta-Analyses Increases the Genomic Prediction Accuracy of Growth and Carcass Traits in Large White Pigs.

Preselected variants associated with the trait of interest from genome-wide association studies (GWASs) are available to improve genomic prediction in pigs. The objectives of this study were to use preselected variants from a large GWAS meta-analysis to assess the impact of single-nucleotide polymorphism (SNP) preselection strategies on genome prediction of growth and carcass traits in pigs. We genotyped 1018 Large White pigs using medium (50k) SNP arrays and then imputed SNPs to sequence level by utilizing a reference panel of 1602 whole-genome sequencing samples. We tested the effects of different proportions of selected top SNPs across different SNP preselection strategies on genomic prediction. Finally, we compared the prediction accuracies by employing genomic best linear unbiased prediction (GBLUP), genomic feature BLUP and three weighted GBLUP models. SNP preselection strategies showed an average improvement in accuracy ranging from 0.3 to 2% in comparison to the SNP chip data. The accuracy of genomic prediction exhibited a pattern of initial increase followed by decrease, or continuous decrease across various SNP preselection strategies, as the proportion of selected top SNPs increased. The highest level of prediction accuracy was observed when utilizing 1 or 5% of top SNPs. Compared with the GBLUP model, the utilization of estimated marker effects from a GWAS meta-analysis as SNP weights in the BLUP|GA model improved the accuracy of genomic prediction in different SNP preselection strategies. The new SNP preselection strategies gained from this study bring opportunities for genomic prediction in limited-size populations in pigs.

Prescreening of large-effect markers with multiple strategies improves the accuracy of genomic prediction

Presently, integrating multi-omics information into a prediction model has become a ameliorate strategy for genomic selection to improve genomic prediction accuracy. Here, we set the genomic and transcriptomic data as the training population data, using BSLMM, TWAS, and eQTL mapping to prescreen features according to | ^βb|>0, top 1% of phenotypic variation explained (PVE), expression-associated single nucleotide polymorphisms (eSNPs), and egenes (false discovery rate (FDR)<0.01), where these loci were set as extra fixed effects (named GBLUP-Fix) and random effects (GFBLUP) to improve the prediction accuracy in the validation population, respectively. The results suggested that both GBLUP-Fix and GFBLUP models could improve the accuracy of longissimus dorsi muscle (LDM), water holding capacity (WHC), shear force (SF), and pH in Huaxi cattle on average from 2.14 to 8.69%, especially the improvement of GFBLUP-TWAS over GBLUP was 13.66% for SF. These methods also captured more genetic variance than GBLUP. Our study confirmed that multi-omics-assisted large-effects loci prescreening could improve the accuracy of genomic prediction.

Prioritized imputed sequence variants from multi-population GWAS improve prediction accuracy for sea lice count in Atlantic salmon (Salmo salar)

Sea lice infestation is one of the major fish health problems that occurs during the grow-out phase in Atlantic salmon (Salmo salar) aquaculture. In this study, we integrated different genomic approaches, including whole-genome sequencing (WGS), genotype imputation and meta-analysis of genome-wide association studies (GWAS), to identify single-nucleotide polymorphisms (SNPs) associated with sea lice count in Atlantic salmon. Different sets of trait-associated SNPs were prioritized and compared against randomly chosen markers, based on the accuracy of genomic predictions for the trait. Lice count phenotypes and dense genotypes of five breeding populations challenged with sea lice were used. Genotype imputation was applied to increase SNP density of challenged animals to WGS level. The summary statistics from GWAS of each population were then combined in a meta-analysis to increase the sample size and improve the statistical power of associations. Eight different genotyping scenarios were considered for genomic prediction: 70K_array: 70 K standard genotyping panel; 70K_priori: 70 K SNPs with the highest p-values identified in the meta-analysis; 30K_priori: 30 K SNPs with the highest p-values identified in the meta-analysis; WGS: SNPs imputed to whole-genome sequencing level. The remaining four scenarios were the same SNP sets with a linkage disequilibrium (LD) pruning filter: 70K_array_LD; 70K_priori_LD; 30K_priori_LD and WGS_LD. Genomic prediction accuracy was evaluated using a five-fold cross-validation scheme in two different populations which were excluded from the meta-analysis to remove possible validation-reference bias. Results showed significant genetic variation for the number of sea lice in Atlantic salmon across populations, with heritabilities ranging from 0.06 to 0.24. The meta-analysis identified several SNPs associated with sea lice resistance, mainly in Ssa03 and Ssa09 chromosomes. Genomic prediction using the GWAS-based prioritized SNPs showed higher accuracy compared to the standard SNP array in most of scenarios; achieving up to 57% increase in accuracy. Accuracy of prioritized scenarios was higher for the 70K_priori than the 30K_priori. The use of WGS data in genomic prediction presented marginal or negative accuracy gain compared to the standard SNP array. The LD-pruning filter presented no benefits, reducing accuracy in most scenarios. Overall, our study demonstrated the potential of prioritized imputed sequence variants from multi-population GWAS meta-analysis to improve prediction accuracy for sea lice counts in Atlantic salmon. The findings suggest that incorporating WGS data and prioritized SNPs from GWAS meta-analysis can accelerate the genetic progress of selection for polygenic traits in salmon aquaculture.

Assessing the Potential for Genome-Assisted Breeding in Red Perilla Using Quantitative Trait Locus Analysis and Genomic Prediction.

Red perilla is an important medicinal plant used in Kampo medicine. The development of elite varieties of this species is urgently required. Medicinal compounds are generally considered target traits in medicinal plant breeding; however, selection based on compound phenotypes (i.e., conventional selection) is expensive and time consuming. Here, we propose genomic selection (GS) and marker-assisted selection (MAS), which use marker information for selection, as suitable selection methods for medicinal plants, and we evaluate the effectiveness of these methods in perilla breeding. Three breeding populations generated from crosses between one red and three green perilla genotypes were used to elucidate the genetic mechanisms underlying the production of major medicinal compounds using quantitative trait locus analysis and evaluating the accuracy of genomic prediction (GP). We found that GP had a sufficiently high accuracy for all traits, confirming that GS is an effective method for perilla breeding. Moreover, the three populations showed varying degrees of segregation, suggesting that using these populations in breeding may simultaneously enhance multiple target traits. This study contributes to research on the genetic mechanisms of the major medicinal compounds of red perilla, as well as the breeding efficiency of this medicinal plant.

Accelerating genetic gains for quantitative resistance to verticillium wilt through predictive breeding in strawberry.

Verticillium wilt (VW), a devastating vascular wilt disease of strawberry (Fragaria ananassa), has caused economic losses for nearly a century. This disease is caused by the soil-borne pathogen Verticillium dahliae, which occurs nearly worldwide and causes disease in numerous agriculturally important plants. The development of VW-resistant cultivars is critically important for the sustainability of strawberry production. We previously showed that a preponderance of the genetic resources (asexually propagated hybrid individuals) preserved in public germplasm collections were moderately to highly susceptible and that genetic gains for increased resistance to VW have been negligible over the last 60 years. To more fully understand the challenges associated with breeding for increased quantitative resistance to this pathogen, we developed and phenotyped a training population of hybrids ( ) among elite parents with a wide range of resistance phenotypes. When these data were combined with training data from a population of elite and exotic hybrids ( ), genomic prediction accuracies of 0.47-0.48 were achieved and were predicted to explain 70%-75% of the additive genetic variance for resistance. We concluded that breeding values for resistance to VW can be predicted with sufficient accuracy for effective genomic selection with routine updating of training populations.

Exploring the potential of incremental feature selection to improve genomic prediction accuracy

BackgroundThe ever-increasing availability of high-density genomic markers in the form of single nucleotide polymorphisms (SNPs) enables genomic prediction, i.e. the inference of phenotypes based solely on genomic data, in the field of animal and plant breeding, where it has become an important tool. However, given the limited number of individuals, the abundance of variables (SNPs) can reduce the accuracy of prediction models due to overfitting or irrelevant SNPs. Feature selection can help to reduce the number of irrelevant SNPs and increase the model performance. In this study, we investigated an incremental feature selection approach based on ranking the SNPs according to the results of a genome-wide association study that we combined with random forest as a prediction model, and we applied it on several animal and plant datasets.ResultsApplying our approach to different datasets yielded a wide range of outcomes, i.e. from a substantial increase in prediction accuracy in a few cases to minor improvements when only a fraction of the available SNPs were used. Compared with models using all available SNPs, our approach was able to achieve comparable performances with a considerably reduced number of SNPs in several cases. Our approach showcased state-of-the-art efficiency and performance while having a faster computation time.ConclusionsThe results of our study suggest that our incremental feature selection approach has the potential to improve prediction accuracy substantially. However, this gain seems to depend on the genomic data used. Even for datasets where the number of markers is smaller than the number of individuals, feature selection may still increase the performance of the genomic prediction. Our approach is implemented in R and is available at https://github.com/FelixHeinrich/GP_with_IFS/.

PSVII-3 The Impact of Genotyping Strategies on Accuracy of Genomic Prediction in Sheep Populations: Preliminary Results of a Simulation Study

Abstract Genomic predictions provide more accurate estimated breeding values (EBV) in younger animals, enabling greater rates of genetic gains over time. However, genomic selection in U.S. sheep is still incipient due to the higher genotyping cost relative to animal cost. Our objective was to compare genotyping strategies differing in the proportion of animals genotyped and the extent of pedigree error on prediction accuracy. These genetic evaluations were based on single-step Genomic Best Linear Unbiased Prediction. We used the AlphaSimR package to simulate a composite sheep population mimicking the Katahdin breeds’ formation. We then simulated four further generations based on the structure of Katahdin flocks engaged in the National Sheep Improvement Program. Thirty-five flocks were simulated with an effective population size of 170. Flock size ranged from 31 to 680 animals and, by design, was skewed towards smaller flocks. Average ram to ewe ratio, mortality rate per generation, and number of lambs born per ewe lambing was 1:16, 10%, and 1.8 lambs, respectively. We simulated a 50k SNP panel for a moderately heritable (0.30) trait. Flocks had low genetic connectedness. Selection criteria differed among flocks: 3 selected randomly, 10 selected based on highest EBV, and 22 selected based on highest phenotypes. All combinations of 5, 40, or 100% of males or females genotyped, 5 or 20% of animals missing pedigree, and 5 or 20% of sires misidentified were compared. We also considered a scenario with no pedigree errors (i.e., no missing pedigree or misidentified sires). Animals with the best phenotype were chosen for genotyping. The first three generations were used as the reference population (7,317 animals) to validate the 4th generation (2,055 animals). Each scenario was replicated 5 times. A linear model was used to estimate differences in prediction accuracy between scenarios, tested with Tukey’s test (alpha 0.05). For the scenario without pedigree errors, prediction accuracy ranged from 0.50 ± 0.06 for 5% of males and females genotyped to 0.75 ± 0.03 for 100% of animals genotyped (Figure 1). For scenarios with 5% pedigree errors, prediction accuracy ranged from 0.33 ± 0.02 for 5% of males and females genotyped to 0.71 ± 0.02 for 100% of animals genotyped. For the scenario with 20% pedigree errors, prediction accuracy ranged from 0.26 ± 0.06 for 5% of males and females genotyped to 0.67 ± 0.03 for 100% of animals genotyped. Across pedigree error scenarios, trends were similar. Importantly, genotyping 5% of one sex and 100% of the other resulted in similar accuracies to genotyping 100% of animals. In scenarios with few animals genotyped, accuracies were higher when errors in pedigree were avoided. These preliminary results reinforce the importance of genomics in correcting pedigree errors, with high levels of prediction accuracy when at least most animals of one sex are genotyped.