Genomic prediction for sugarcane diseases including hybrid Bayesian-machine learning approaches.

Sugarcane smut and Pachymetra root rots are two serious diseases of sugarcane, with susceptible infected crops losing over 30% of yield. A heritable component to both diseases has been demonstrated, suggesting selection could improve disease resistance. Genomic selection could accelerate gains even further, enabling early selection of resistant seedlings for breeding and clonal propagation. In this study we evaluated four types of algorithms for genomic predictions of clonal performance for disease resistance. These algorithms were: Genomic best linear unbiased prediction (GBLUP), including extensions to model dominance and epistasis, Bayesian methods including BayesC and BayesR, Machine learning methods including random forest, multilayer perceptron (MLP), modified convolutional neural network (CNN) and attention networks designed to capture epistasis across the genome-wide markers. Simple hybrid methods, that first used BayesR/GWAS to identify a subset of 1000 markers with moderate to large marginal additive effects, then used attention networks to derive predictions from these effects and their interactions, were also developed and evaluated. The hypothesis for this approach was that using a subset of markers more likely to have an effect would enable better estimation of interaction effects than when there were an extremely large number of possible interactions, especially with our limited data set size. To evaluate the methods, we applied both random five-fold cross-validation and a structured PCA based cross-validation that separated 4702 sugarcane clones (that had disease phenotypes and genotyped for 26k genome wide SNP markers) by genomic relationship. The Bayesian methods (BayesR and BayesC) gave the highest accuracy of prediction, followed closely by hybrid methods with attention networks. The hybrid methods with attention networks gave the lowest variation in accuracy of prediction across validation folds (and lowest MSE), which may be a criteria worth considering in practical breeding programs. This suggests that hybrid methods incorporating the attention mechanism could be useful for genomic prediction of clonal performance, particularly where non-additive effects may be important.

Use of continuous genotypes for genomic prediction in sugarcane.

Genomic selection in sugarcane faces challenges due to limited genomic tools and high genomic complexity, particularly because of its high and variable ploidy. The classification of genotypes for single nucleotide polymorphisms (SNPs) becomes difficult due to the wide range of possible allele dosages. Previous genomic studies in sugarcane used pseudo-diploid genotyping, grouping all heterozygotes into a single class. In this study, we investigate the use of continuous genotypes as a proxy for allele-dosage in genomic prediction models. The hypothesis is that continuous genotypes could better reflect allele dosage at SNPs linked to mutations affecting target traits, resulting in phenotypic variation. The dataset included genotypes of 1318 clones at 58K SNP markers, with about 26K markers filtered using standard quality controls. Predictions for tonnes of cane per hectare (TCH), commercial cane sugar (CCS), and fiber content (Fiber) were made using parametric, non-parametric, and Bayesian methods. Continuous genotypes increased accuracy by 5%-7% for CCS and Fiber. The pseudo-diploid parametrization performed better for TCH. Reproducing kernel Hilbert spaces model with Gaussian kernel and AK4 (arc-cosine kernel with hidden layer 4) kernel outperformed other methods for TCH and CCS, suggesting that non-additive effects might influence these traits. The prevalence of low-dosage markers in the study may have limited the benefits of approximating allele-dosage information with continuous genotypes in genomic prediction models. Continuous genotypes simplify genomic prediction in polyploid crops, allowing additional markers to be used without adhering to pseudo-diploid inheritance. The approach can particularly benefit high ploidy species or emerging crops with unknown ploidy.

Optimising clonal performance in sugarcane: leveraging non-additive effects via mate-allocation strategies.

Mate-allocation strategies in breeding programs can improve progeny performance by harnessing non-additive genetic effects. These approaches prioritise predicted progeny merit over parental breeding value, making them particularly appealing for clonally propagated crops such as sugarcane. We conducted a comparative analysis of mate-allocation strategies, exploring utilising non-additive and heterozygosity effects to maximise clonal performance with schemes that solely consider additive effects to optimise breeding value. Using phenotypic and genotypic data from a population of 2,909 clones evaluated in final assessment trials of Australian sugarcane breeding programs, we focused on three important traits: tonnes of cane per hectare (TCH), commercial cane sugar (CCS), and Fibre. By simulating families from all possible crosses (1,225) with 50 progenies each, we predicted the breeding and clonal values of progeny using two models: GBLUP (considering additive effects only) and extended-GBLUP (incorporating additive, non-additive, and heterozygosity effects). Integer linear programming was used to identify the optimal mate-allocation among selected parents. Compared to breeding value-based approaches, mate-allocation strategies based on clonal performance yielded substantial improvements, with predicted progeny values increasing by 57% for TCH, 12% for CCS, and 16% for fibre. Our simulation study highlights the effectiveness of mate-allocation approaches that exploit non-additive and heterozygosity effects, resulting in superior clonal performance. However, there was a notable decline in additive gain, particularly for TCH, likely due to significant epistatic effects. When selecting crosses based on clonal performance for TCH, the inbreeding coefficient of progeny was significantly lower compared to random mating, underscoring the advantages of leveraging non-additive and heterozygosity effects in mitigating inbreeding depression. Thus, mate-allocation strategies are recommended in clonally propagated crops to enhance clonal performance and reduce the negative impacts of inbreeding.

Genomic prediction with machine learning in sugarcane, a complex highly polyploid clonally propagated crop with substantial non-additive variation for key traits.

Sugarcane has a complex, highly polyploid genome with multi-species ancestry. Additive models for genomic prediction of clonal performance might not capture interactions between genes and alleles from different ploidies and ancestral species. As such, genomic prediction in sugarcane presents an interesting case for machine learning (ML) methods, which are purportedly able to deal with high levels of complexity in prediction. Here, we investigated deep learning (DL) neural networks, including multilayer networks (MLP) and convolution neural networks (CNN), and an ensemble machine learning approach, random forest (RF), for genomic prediction in sugarcane. The data set used was 2912 sugarcane clones, scored for 26,086 genome wide single nucleotide polymorphism markers, with final assessment trial data for total cane harvested (TCH), commercial cane sugar (CCS), and fiber content (Fiber). The clones in the latest trial (2017) were used as a validation set. We compared prediction accuracy of these methods to genomic best linear unbiased prediction (GBLUP) extended to include dominance and epistatic effects. The prediction accuracies from GBLUP models were up to 0.37 for TCH, 0.43 for CCS, and 0.48 for Fiber, while the optimized ML models had prediction accuracies of 0.35 for TCH, 0.38 for CCS, and 0.48 for Fiber. Both RF and DL neural network models have comparable predictive ability with the additive GBLUP model but are less accurate than the extended GBLUP model.

Breed identification using breed-informative SNPs and machine learning based on whole genome sequence data and SNP chip data.

BACKGROUND: Breed identification is useful in a variety of biological contexts. Breed identification usually involves two stages, i.e., detection of breed-informative SNPs and breed assignment. For both stages, there are several methods proposed. However, what is the optimal combination of these methods remain unclear. In this study, using the whole genome sequence data available for 13 cattle breeds from Run 8 of the 1,000 Bull Genomes Project, we compared the combinations of three methods (Delta, FST, and In) for breed-informative SNP detection and five machine learning methods (KNN, SVM, RF, NB, and ANN) for breed assignment with respect to different reference population sizes and difference numbers of most breed-informative SNPs. In addition, we evaluated the accuracy of breed identification using SNP chip data of different densities. RESULTS: We found that all combinations performed quite well with identification accuracies over 95% in all scenarios. However, there was no combination which performed the best and robust across all scenarios. We proposed to integrate the three breed-informative detection methods, named DFI, and integrate the three machine learning methods, KNN, SVM, and RF, named KSR. We found that the combination of these two integrated methods outperformed the other combinations with accuracies over 99% in most cases and was very robust in all scenarios. The accuracies from using SNP chip data were only slightly lower than that from using sequence data in most cases. CONCLUSIONS: The current study showed that the combination of DFI and KSR was the optimal strategy. Using sequence data resulted in higher accuracies than using chip data in most cases. However, the differences were generally small. In view of the cost of genotyping, using chip data is also a good option for breed identification.

Combining genomic selection with genome-wide association analysis identified a large-effect QTL and improved selection for red rot resistance in sugarcane.

Red rot caused by the fungus Colletotrichum falcatum is the main disease limiting sugarcane productivity in several countries including the major producer India. The genetic basis for red rot resistance is unclear. We studied a panel of 305 sugarcane clones from the Australian breeding program for disease response phenotype and genotype using an Affymetrix® Axiom® array, to better understand the genetic basis of red rot resistance. SNP markers highly significantly associated with red rot response (≤ 10-8) were identified. Markers with largest effect were located in a single 14.6 Mb genomic region of sorghum (the closest diploid relative of sugarcane with a sequenced genome) suggesting the presence of a major-effect QTL. By genomic selection, the estimated selection accuracy was ~0.42 for red rot resistance. This was increased to ~0.5 with the addition of 29 highly significant SNPs as fixed effects. Analysis of genes nearby the markers linked to the QTL revealed many biotic stress responsive genes within this QTL, with the most significant SNP co-locating with a cluster of four chitinase A genes. The SNP markers identified here could be used to predict red rot resistance with high accuracy at any stage in the sugarcane breeding program.

A linkage disequilibrium-based approach to position unmapped SNPs in crop species.

BACKGROUND: High-density SNP arrays are now available for a wide range of crop species. Despite the development of many tools for generating genetic maps, the genome position of many SNPs from these arrays is unknown. Here we propose a linkage disequilibrium (LD)-based algorithm to allocate unassigned SNPs to chromosome regions from sparse genetic maps. This algorithm was tested on sugarcane, wheat, and barley data sets. We calculated the algorithm's efficiency by masking SNPs with known locations, then assigning their position to the map with the algorithm, and finally comparing the assigned and true positions. RESULTS: In the 20-fold cross-validation, the mean proportion of masked mapped SNPs that were placed by the algorithm to a chromosome was 89.53, 94.25, and 97.23% for sugarcane, wheat, and barley, respectively. Of the markers that were placed in the genome, 98.73, 96.45 and 98.53% of the SNPs were positioned on the correct chromosome. The mean correlations between known and new estimated SNP positions were 0.97, 0.98, and 0.97 for sugarcane, wheat, and barley. The LD-based algorithm was used to assign 5920 out of 21,251 unpositioned markers to the current Q208 sugarcane genetic map, representing the highest density genetic map for this species to date. CONCLUSIONS: Our LD-based approach can be used to accurately assign unpositioned SNPs to existing genetic maps, improving genome-wide association studies and genomic prediction in crop species with fragmented and incomplete genome assemblies. This approach will facilitate genomic-assisted breeding for many orphan crops that lack genetic and genomic resources.

Improved genomic prediction of clonal performance in sugarcane by exploiting non-additive genetic effects.

KEY MESSAGE: Non-additive genetic effects seem to play a substantial role in the expression of complex traits in sugarcane. Including non-additive effects in genomic prediction models significantly improves the prediction accuracy of clonal performance. In the recent decade, genetic progress has been slow in sugarcane. One reason might be that non-additive genetic effects contribute substantially to complex traits. Dense marker information provides the opportunity to exploit non-additive effects in genomic prediction. In this study, a series of genomic best linear unbiased prediction (GBLUP) models that account for additive and non-additive effects were assessed to improve the accuracy of clonal prediction. The reproducible kernel Hilbert space model, which captures non-additive genetic effects, was also tested. The models were compared using 3,006 genotyped elite clones measured for cane per hectare (TCH), commercial cane sugar (CCS), and Fibre content. Three forward prediction scenarios were considered to investigate the robustness of genomic prediction. By using a pseudo-diploid parameterization, we found significant non-additive effects that accounted for almost two-thirds of the total genetic variance for TCH. Average heterozygosity also had a major impact on TCH, indicating that directional dominance may be an important source of phenotypic variation for this trait. The extended-GBLUP model improved the prediction accuracies by at least 17% for TCH, but no improvement was observed for CCS and Fibre. Our results imply that non-additive genetic variance is important for complex traits in sugarcane, although further work is required to better understand the variance component partitioning in a highly polyploid context. Genomics-based breeding will likely benefit from exploiting non-additive genetic effects, especially in designing crossing schemes. These findings can help to improve clonal prediction, enabling a more accurate identification of variety candidates for the sugarcane industry.

Accuracy of genomic prediction of complex traits in sugarcane.

KEY MESSAGE: Complex traits in sugarcane can be accurately predicted using genome-wide DNA markers. Genomic single-step prediction is an attractive method for genomic selection in commercial breeding programs. Sugarcane breeding programs have achieved up to 1% genetic gain in key traits such as tonnes of cane per hectare (TCH), commercial cane sugar (CCS) and Fibre content over the past decades. Here, we assess the potential of genomic selection to increase the rate of genetic gain for these traits by deriving genomic estimated breeding values (GEBVs) from a reference population of 3984 clones genotyped for 26 K SNP. We evaluated the three different genomic prediction approaches GBLUP, genomic single step (GenomicSS), and BayesR. GenomicSS combining pedigree and SNP information from historic and recent breeding programs achieved the most accurate predictions for most traits (0.3-0.44). This method is attractive for routine genetic evaluation because it requires relatively little modification to the existing evaluation and results in breeding value estimates for all individuals, not only those genotyped. Adding information from early-stage trials added up to 5% accuracy for CCS and Fibre, but 0% for TCH, reflecting the importance of competition effects for TCH. These GEBV accuracies are sufficiently high that, combined with the right breeding strategy, a doubling of the rate of genetic gain could be achieved. We also assessed the flowering traits days to flowering, gender and pollen viability and found high heritabilities of 0.57, 0.78 and 0.72, respectively. The GEBV accuracies indicated that genomic selection could be used to improve these traits. This could open new avenues for breeders to manage their breeding programs, for example, by synchronising flowering time and selecting males with high pollen viability.

Strategies and considerations for implementing genomic selection to improve traits with additive and non-additive genetic architectures in sugarcane breeding.

KEY MESSAGE: Simulations highlight the potential of genomic selection to substantially increase genetic gain for complex traits in sugarcane. The success rate depends on the trait genetic architecture and the implementation strategy. Genomic selection (GS) has the potential to increase the rate of genetic gain in sugarcane beyond the levels achieved by conventional phenotypic selection (PS). To assess different implementation strategies, we simulated two different GS-based breeding strategies and compared genetic gain and genetic variance over five breeding cycles to standard PS. GS scheme 1 followed similar routines like conventional PS but included three rapid recurrent genomic selection (RRGS) steps. GS scheme 2 also included three RRGS steps but did not include a progeny assessment stage and therefore differed more fundamentally from PS. Under an additive trait model, both simulated GS schemes achieved annual genetic gains of 2.6-2.7% which were 1.9 times higher compared to standard phenotypic selection (1.4%). For a complex non-additive trait model, the expected annual rates of genetic gain were lower for all breeding schemes; however, the rates for the GS schemes (1.5-1.6%) were still greater than PS (1.1%). Investigating cost-benefit ratios with regard to numbers of genotyped clones showed that substantial benefits could be achieved when only 1500 clones were genotyped per 10-year breeding cycle for the additive genetic model. Our results show that under a complex non-additive genetic model, the success rate of GS depends on the implementation strategy, the number of genotyped clones and the stage of the breeding program, likely reflecting how changes in QTL allele frequencies change additive genetic variance and therefore the efficiency of selection. These results are encouraging and motivate further work to facilitate the adoption of GS in sugarcane breeding.

Simultaneously accounting for population structure, genotype by environment interaction, and spatial variation in marker-trait associations in sugarcane.

Few association mapping studies have simultaneously accounted for population structure, genotype by environment interaction (GEI), and spatial variation. In this sugarcane association mapping study we tested models accounting for these factors and identified the impact that each model component had on the list of markers declared as being significantly associated with traits. About 480 genotypes were evaluated for cane yield and sugar content at three sites and scored with DArT markers. A mixed model was applied in analysis of the data to simultaneously account for the impacts of population structure, GEI, and spatial variation within a trial. Two forms of the DArT marker data were used in the analysis: the standard discrete data (0, 1) and a continuous DArT score, which is related to the marker dosage. A large number of markers were significantly associated with cane yield and sugar content. However, failure to account for population structure, GEI, and (or) spatial variation produced both type I and type II errors, which on the one hand substantially inflated the number of significant markers identified (especially true for failing to account for GEI) and on the other hand resulted in failure to detect markers that could be associated with cane yield or sugar content (especially when failing to account for population structure). We concluded that association mapping based on trials from one site or analysis that failed to account for GEI would produce many trial-specific associated markers that would have low value in breeding programs.

Joint modeling of additive and non-additive (genetic line) effects in multi-environment trials.

A statistical approach for the analysis of multi-environment trials (METs) is presented, in which selection of best performing lines, best parents, and best combination of parents can be determined. The genetic effect of a line is partitioned into additive, dominance and residual non-additive effects. The dominance effects are estimated through the incorporation of the dominance relationship matrix, which is presented under varying levels of inbreeding. A computationally efficient way of fitting dominance effects is presented which partitions dominance effects into between family dominance and within family dominance line effects. The overall approach is applicable to inbred lines, hybrid lines and other general population structures where pedigree information is available.

Associations between DNA markers and resistance to diseases in sugarcane and effects of population substructure.

Association between markers and sugarcane diseases were investigated in a collection of 154 sugarcane clones, consisting of important ancestors or parents, and cultivars. 1,068 polymorphic AFLP and 141 SRR markers were scored across all clones. Data on the four most important diseases in the Australian sugarcane industry were obtained; these diseases being pachymetra root rot (Pachymetra chaunorhiza B.J. Croft & M.W. Dick), leaf scald (Xanthomonas albilineans Dowson), Fiji leaf gall (Fiji disease virus), and smut (Ustilago scitaminea H. & P. Sydow). By a simple regression analysis, association between markers and diseases could be readily detected. However, many of these associations were due to the effects of embedded population structure and random effects. After taking population structure into account, we found that 59% of the phenotypic variation in smut resistance ratings could be accounted for by 11 markers, 32% of variation for leaf scald and pachymetra root rot rating by 4 markers, and 26% of Fiji leaf gall by 5 markers. The results suggest that marker-trait associations can be readily detected in populations generated from modern sugarcane breeding programs. This may be due to special features of past sugarcane breeding programs leading to persistent linkage disequilibrium in modern parental populations.