Modelos mistos na seleção entre e dentro de famílias de cana de açúcar sob o enfoque bayesiano

• Ano de defesa: 2012
• Autor(a) principal: Silva, Mariane Alves Gomes da
• Orientador(a): Peternelli, Luiz Alexandre
• Banca de defesa: Nascimento, Moysés , Silva, Felipe Lopes da , Martins Filho, Sebastião
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Estatística Aplicada e Biometria
• Departamento: Estatística Aplicada e Biometria
• País: BR
• Palavras-chave em Português: Análise Bayesiana; Melhoramento genético; BLUPIS
• Palavras-chave em Inglês: Bayesian analysis; Genetic improvement; BLUPIS
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA::ESTATISTICA
• Link de acesso: http://locus.ufv.br/handle/123456789/4050

The base of agribusiness of sugar cane is the breeding. Can be shown that the great strategy selection of plant would be through the prediction of genotypic values using BLUP (Best Linear Unbiased Predictor) individual (BLUPI). This procedure would use both the information of family and individuals for selection. However this method is rarely used in breeding programs because of operational problems related to obtain data plant. Recently an alternative operationally more practical was proposal and is called BLUPIS (BLUP individual simulated). In this case the data are collected at the level of plot. With this is possible to select the best families, and subsequently, to simulate the number of individuals to be selected within the best families. This work has as aims to develop an algorithm for analysis BLUP under the Bayesian focus with different settings of priors in your modeling, in statistical software R, for possible available to the user and compare it with the Classic REML / BLUP. The results showed that the method with Bayesian BLUPIS performed by the algorithm built with the R program was effective. The algorithm took into account the uncertainties on all parameters, as also allowed the use of a priori information. The Bayesian method is more efficient when considering the relationship information model and the distribution of the prior information (major genotypic effects and variances and heritability smaller).