Optimizing GWS and GWAS in crop breeding

• Ano de defesa: 2021
• Autor(a) principal: Torres, Lívia Gomes
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Mandioca - Melhoramento genético; Soja - Melhoramento genético; Genômica; Predição; Mapeamento genômico vegetal; Genética Quantitativa
• Link de acesso: https://locus.ufv.br//handle/123456789/28762

In a scenario of advances in genotyping technologies, with the decreasing cost per sample, the information obtained at the molecular level has been increasingly used in genetic analyses to accelerate the effective selection of clones, through genomic prediction (GP). GP leads to more accurate predictions of the individual’s merit for the traits of interest. Genome- wide association studies (GWAS) have also been widely used, it aims to identify causal variants to detect significant associations between genetic markers and traits of interest. In the first chapter, predictions based on pedigree information, molecular markers and cassava productive traits were done with clones from biparental crosses by separating the full dataset by the number of stages of a breeding program (clonal evaluation trial – CET, preliminary yield trial – PYT, advanced yield trial – AYT, and uniform yield trial – UYT): one stage (CET), two stages (CET and PYT), three stages (CET, PYT and AYT) and four stages (CET, PYT, AYT and UYT). The results indicated a satisfactory potential for using genomic prediction in cassava, especially for early selection, thus saving resources. The second chapter was a case study with hydrogen cyanide content (HCN) in cassava, to evaluate the feasibility of using cross-country genomic predictions, with datasets from Brazil and Nigeria, as a strategy to assist clones’ selection in germplasm exchange. In addition, it also provided an assessment of the population structure for the joint dataset with the two countries as well as genetic parameters estimations based on single nucleotide polymorphisms (SNPs) and in a haplotype approach. Comparisons on GEBVs’ estimation were made considering the hypothetical situation of not having the phenotypic characterization for a set of clones for a certain research institute/country and might need to use the markers’ effects that were trained with data from other research institute/country’s germplasm to estimate their clones’ GEBV. The joint dataset provided an improved accuracy compared to the prediction accuracy of either germplasm’ sources individually. Cross-country genomic predictions proved to have potential use under the present study’s scenario, the correlation between GEBVs predicted with TP from Embrapa and IITA was 0.55 for Embrapa’s germplasm, whereas for IITA’s it was 0.10. In the third chapter, a case study was proposed as well, it was related to association mapping in soybean breeding, for protein and oil content. The benefits of increasing the marker coverage in GWAS analyses were investigated. This study aimed to explore this issue by comparing the results of a genome-wide association study (GWAS) for seed oil and protein content performed with either 17K GBS-derived SNPs or 2.18M GBS-/WGS-derived SNPs on the same set of soybean accessions. Our results revealed that comparatively to GWAS with WGS-imputed SNPs, GWAS analyses with a very limited number of markers have been successful in identifying relevant regions associated with protein and oil content in soybean. The results of this chapter also provided new insights into the application of imputation with WGS data in soybean breeding. These chapters sought to address issues related to the choice of training population, through stages of a breeding program, as well as between breeding programs (from different countries) in order to either have a better understanding of the possibility of shortening the long cassava breeding cycle, as well as of the use of genomic information to guide germplasm’s exchange when there is no phenotypic information for clones from a given research institute/country. In the genome association analyses for soybeans, the objective was to verify the extent to which the work and extra cost involved with WGS-imputation is worthy, because, for some researchers, the level of resolution required for marker-assisted selection would already be supplied by GBS, but that would certainly depend on the research objectives, species and trait heritability. Keywords: Manihot esculenta. Genomic prediction. Training population. genotyping-by- sequencing. Soybean. GWAS. Association mapping.