Métodos de estimação e validação na seleção genômica na presença ou ausência de correção de fenótipos

• Ano de defesa: 2013
• Autor(a) principal: Almeida, ísis Fernanda de
• Orientador(a): Cruz, Cosme Damião
• Banca de defesa: Bhering, Leonardo Lopes , Oliveira, Marciane da Silva , Almeida, Ramon Vinicius de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Doutorado em Genética e Melhoramento
• Departamento: Genética animal; Genética molecular e de microrganismos; Genética quantitativa; Genética vegetal; Me
• País: BR
• Palavras-chave em Português: Seleção genômica; BLASSO; Desequilíbrio de ligação
• Palavras-chave em Inglês: Genomic selection; BLASSO; Linkage disequilibrium
• Área do conhecimento CNPq: CNPQ::CIENCIAS BIOLOGICAS::GENETICA::GENETICA QUANTITATIVA
• Link de acesso: http://locus.ufv.br/handle/123456789/1386

With new breakthroughs in large-scale genotyping technologies, it became possible to achieve a genome complete coverage, which led the researchers to create a new way of using this new genetic information acquired to the benefit of plant breeding. This new methodology was called genome-wide selection (GWS), and consists of a simultaneous use of hundreds or thousands of markers, which have a wide and dense coverage of the genome then all genes of a quantitative trait are in linkage disequilibrium with at least one of those markers. Thus, that study aimed to evaluate the accuracy of genomic selection in different scenarios of genetic distribution effects and validation populations, and analyze the effect of phenotype adjustment in the population structure. Once the data were generated by simulation, it was also observed the effectiveness of that simulation process to generate populations whose genetic principles are preserved. For this, ten replicates of two population structures were simulated, with 500 individuals each. For genotypic data was considered 1000 marker loci with 100 effectively linked to QTL. Distinct QTL effects were simulated, one according to a uniform distribution and another according to an exponential distribution with the goal of represent different genetic architectures. For validation 1 (V1) a sample of 100 individuals composed the validation population, for validation 2 (V2), we applied the Jacknife cross-validation, and validation 3 (V3), a new validation population was generated. The RR-BLUP and BLASSO methodologies were used in order to compute the genomic selection accuracy on the different established scenarios. The simulation of populations was effective in represent the genetic structure of populations. Without phenotypic correction for population structure the QTL exponential distribution effects led to higher accuracy, the BLASSO method showed more accuracy for this distribution of genetic effects. In this scenario, the validations V1 and V3 were more accurated. With phenotypes adjustement the QTL genetic effects in exponential and uniform distributions led to similar accuracies and BLASSO method proved more robust and more accurate for both distributions of genetic effects. In this situation, validations V1 and V2 showed up more accurated. The phenotypes adjustement led to higher accuracy due to greater efficiency in capturing linkage disequilibrium. Generally the RR-BLUP method was more accurate and the BLASSO method less biased.