Influência do tamanho amostral e do tipo de amostragem na estimação dos efeitos de marcadores SNPs e predição de valores genéticos genômicos

• Autor(a) principal: Sousa, Mariele Freitas
• Data de Publicação: 2013
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/1850
• topic: Avaliação genética; Parâmetros genéticos; Simulação; Seleção genômica; Genetic avaliation; Genetic parameters; Simulation; Genomic selection; CNPQ::CIENCIAS AGRARIAS::ZOOTECNIA::GENETICA E MELHORAMENTO DOS ANIMAIS DOMESTICOS

This work aimed to study the effect of sample size and type of sampling on the estimation of SNPs effects and prediction of genomic breeding values for traits with low, medium and high heritabilities; using the one step methodology. Genotypic data were simulated for 60K SNP loci with frequencies similar to real data. Different types of sampling were compared (random sampling and three generations sampling); sample size (500, 1000 or 2,000 animals) and traits with different heritabilities (0.10, 0.30 and 0.70). The expected breeding values (EBV) were obtained through one step methodology, using BLUPF90 software with default settings. The estimated SNP effects were obtained through one-step methodology using PostGSf90, a BLUPF90 package that converts GEBV into SNP effects. Pearson s correlation was used to represent the accuracy of the estimates of SNP effects and predictions of genomic breeding values. Correlations increased as the genotyped population size and heritability increased. When only SNPs with major effect were considered, high correlations between estimated and simulated SNPs were obtained. Furthermore, it was observed that there is an increase in correlation with the increase in sample size and heritability. For correlation between simulated and estimated genomic breeding values, there was a difference only at heritability levels, where the higher heritability that showed higher correlations. Although using simulated data in complete linkage disequilibrium and one-step approach, it was observed effect of number of genotyped animals and type of sampling, especially on estimation of SNP effects and low heritability traits.