Inclusão de efeitos de dominância no modelo GBLUP multivariado para predição de híbridos simples de milho

• Ano de defesa: 2015
• Autor(a) principal: Santos, Jhonathan Pedroso Rigal dos
• Orientador(a): Von Pinho, Renzo Garcia
• Banca de defesa: Fritsche-Neto, Roberto
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: UNIVERSIDADE FEDERAL DE LAVRAS
• Programa de Pós-Graduação: DBI - Programa de Pós-graduação
• País: BRASIL
• Palavras-chave em Português: Análise univariada; Análise multivariada; Univariate analysis; Multivariate analysis
• Área do conhecimento CNPq: CNPQ_NÃO_INFORMADO
• Link de acesso: https://repositorio.ufla.br/handle/1/9615

New proposals of models and application of prediction processes using molecular markers information can contribute to the reduction of financial expenses and identification of superior genotypes in maize breeding programs. Studies evaluating GBLUP model with the inclusion of dominance effects were not made in the univariate and multivariate context in the analysis of maize data. In this study, it was conducted a construction procedure of simple hybrids with phenotypic data and real molecular markers of 4091 maize lines of the public database Panzea. In this process, 400 simple hybrids were obtained and then analyzed using univariate and multivariate GBLUP model considering only additive effects, and in another configuration with the inclusion of dominance effects. Historical heritability scenario of five maize characters and in other conditions of genetic architecture for comparative research among the models, assessing their predictive capacity and decomposition of variance components. Significant differences were not detected between the multivariate and univariate models. The main reason for this small discrepancy between the models is the low to moderate magnitude correlations among the characters studied and the moderate heritability observed. This condition does not favor the advantages of multivariate analysis. The inclusion of dominance effects in models is an efficient strategy to improve the predictive capacity and the quality of the variance components decomposition.