Introducing Sparsity Into Selection Index Methodology With Applications To High Throughput Phenotyping And Genomic Prediction

Introducing Sparsity Into Selection Index Methodology with Applications to High-throughput Phenotyping and Genomic Prediction

Research in plant and animal breeding has been largely focused on the development of methods for a more efficient selection by altering the factors that affect genetic progress: selection intensity, selection accuracy, genetic variance, and length of the breeding cycle. Most of the breeding efforts have been primarily towards increasing selection accuracy and reducing the breeding cycle.Genomic selection has been successfully adopted by many public and private breeding organizations. Over years, these institutions have developed and accumulated large volumes of genomic data linked to phenotypes from multiple populations and multiple generations. This data abundance offers the opportunity to revolutionize genetic research. However, these data sets are also increasingly heterogeneous, with many subpopulations and multiple generations represented in the data. This translates into potentially heterogeneous allele frequencies and different LD patterns, thus leading to SNP-effect heterogeneity.Genomic selection methods were developed with reference to homogeneous populations in which SNP-effects are assumed constant across the whole population. These methods are not necessarily optimal for the contemporary available data sets for model training. Therefore, a first focus of this dissertation is on developing novel methods that can leverage the large-scale of modern data sets while coping with the heterogeneity and complexity of this type of data.In recent years, there have also been important advances in high-throughput phenotyping (HTP) technologies that can generate large volumes of data at multiple time-points of a crop. Examples of this include hyper-spectral imaging technologies that can capture the reflectance of electromagnetic power by crops at potentially thousands of wavelengths. The integration of HTP in genetic evaluations represents a great opportunity to further advance plant breeding; however, the high-dimensional nature of HTP data poses important challenges. Therefore, a second focus of this dissertation is on the development of a novel approach to efficiently incorporate HTP data for breeding values prediction.Thus, this dissertation aims to contribute novel methods that can improve the accuracy of genomic prediction by optimizing the use of large, potentially heterogeneous, genomic data sets and by enabling the integration of HTP data. We present a novel statistical approach that combines the standard selection index methodology with variable-selection methods commonly used in machine learning and statistics, and developed software to implement the method. Our approach offers solutions to both genomic selection with potentially highly heterogeneous genomic data sets, and the integration of HTP in genetic evaluations.

Artificial Intelligence Applications in Specialty Crops

Genomic Selection: Lessons Learned and Perspectives

Genomic selection (GS) has been the most prominent topic in breeding science in the last two decades. The continued interest is promoted by its huge potential impact on the efficiency of breeding. Predicting a breeding value based on molecular markers and phenotypic values of relatives may be used to manipulate three parameters of the breeder's equation. First, the accuracy of the selection may be improved by predicting the genetic value more reliably when considering the records of relatives and the realized genomic relationship. Secondly, genotyping and predicting may be more cost effective than comprehensive phenotyping. Resources can instead be allocated to increasing population sizes and selection intensity. The third, probably most important factor, is time. As shown in dairy cattle breeding, reducing cycle time by crossing selection candidates earlier may have the strongest impact on selection gain. Many different prediction models have been used, and different ways of using predicted values in a breeding program have been explored. We would like to address the questions: i. How did GS change breeding schemes of different crops in the last 20 years? ii. What was the impact on realized selection gain? iii. What would be the best structure of a crop-specific breeding scheme to exploit the full potential of GS? iv. What is the potential of hybrid prediction, epistasis effect models, deep learning methods and other extensions of the standard prediction of additive effects? v. What are the long-term effects of GS? vi. Can predictive breeding approaches also be used to harness genetic resources from germplasm banks in a more efficient way to adapt current germplasm to new environmental challenges? This Research Topic welcomes submissions of Original Research papers, Opinions, Perspectives, Reviews, and Mini-Reviews related to these themes: 1. Genomic selection: statistical methodology 2. The (optimal) use of GS in breeding schemes 3. Practical experiences with GS (selection gain, long-term effects, negative side effects) 4. Predictive approaches to harness genetic resources Concerning point 1): If an original research paper compares different methods empirically without theoretical considerations on when one or the other method should be better, the methods should be compared with at least five different data sets. The data sets should differ either in crop, genotyping method or its source, for instance from a breeding program or gene bank accessions. Concerning point 2): Manuscripts addressing the use of GS in breeding schemes should illustrate breeding schemes that are run in practice. General ideas about schemes that may be run in the future may be considered as 'Perspective' articles. Conflict of Interest statements: - Topic Editor Valentin Wimmer is affiliated to KWS SAAT SE & Co. KGaA, Germany. - Topic Editor Brian Gardunia is affiliated to Bayer Crop Sciences and has a collaboration with AbacusBio, and is an author on patents with Bayer Crop Sciences. The other Topic Editors did not disclose any conflicts of interest. Image credit: CIMMYT, reproduced under the CC BY-NC-SA 2.0 license