Análise de fatores aplicada em estudos de seleção genômica no melhoramento de Coffea canephora
Arquivos
Data
2020-02-20
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Viçosa
Resumo
O Brasil se destaca em âmbito mundial na produção de café. Os incrementos observados em sua produtividade é resultado do aprimoramento de diversas metodologias. Dentre elas, destacam-se os métodos preditivos de valor genético. Estes contribuem significativamente na seleção de genótipos superiores, de forma a aumentar o ganho genético por unidade de tempo. Neste contexto, a seleção genômica ampla (GWS) é uma ferramenta que se destaca, uma vez que permite predizer o fenótipo futuro de um indivíduo baseado apenas em informações de marcadores moleculares. Realizar a seleção de maneira simultânea para várias características é o interesse da maioria dos programas de melhoramento, e a análise de fatores (AF) tem sido utilizada para auxiliar neste fim. A utilização de fatores se justifica devido a existência de correlações genéticas entre as características, as quais podem ser atribuídas aos QTL que têm efeitos pleiotrópicos ou aos QTL estreitamente ligados. Dessa forma, o objetivo deste trabalho foi de avaliar o uso da AF no contexto de GWS, em genótipos de Coffea canephora. Os resultados obtidos da seleção baseada nos fatores foram comparados, por meio da capacidade preditiva, acurácia e do coeficiente de Cohen’s Kappa, com aqueles advindos da análise das variáveis individuais. Para isso, foram utilizados dados fenotípicos e genotípicos de populações compostas por clones dos grupos varietais Conilon e Robusta e por híbridos originados de cruzamentos entre estes grupos, avaliados durante três anos consecutivos (2014 a 2016), e uma densidade de 18111 marcadores SNPs identificados. A partir dos resultados observados, verificou-se que a AF foi eficiente para elucidar as relações entre as características e originar novas variáveis. Os fatores formados são interessantes em termos de seleção, pois além de permitirem interpretações conjuntas, apresentam boas estimativas de capacidade preditiva, herdabilidade e acurácia. Ademais observou-se alta concordância entre os indivíduos selecionados com base nos fatores e aqueles selecionados considerando as variáveis individuais. Entretanto, cabe destacar que, a seleção baseada nos fatores conseguiu selecionar indivíduos de porte mais adequado. Palavras-chave: Predição Genômica. Análise Multivariada. Melhoramento Genético.
Brazil stands out worldwide in the production of coffee. The increases observed in its productivity are the result of the improvement of several methodologies. Among them, the predictive methods of genetic value stand out. These contribute significantly to the selection of superior genotypes, in order to increase the genetic gain per unit of time. In this context, broad genomic selection (GWS) is a tool that stands out, since it allows predicting the future phenotype of an individual based only on information from molecular markers. Performing the selection simultaneously for various characteristics is the interest of most breeding programs, and factor analysis (AF) has been used to assist in this end. The use of factors is justified due to the existence of genetic correlations between the characteristics, which can be attributed to the QTL that have pleiotropic effects or to the closely linked QTL. Thus, the objective of this work was to evaluate the use of AF in the context of GWS, in genotypes of Coffea canephora. The results obtained from the selection based on the factors were compared, through predictive capacity, accuracy and the Cohen’s Kappa coefficient, with those derived from the analysis of the individual variables. For this, phenotypic and genotypic data from populations composed of clones of the varietal groups Conilon and Robusta and hybrids originated from crosses between these groups, evaluated for three consecutive years (2014 to 2016), and a density of 18111 identified SNPs markers were used. From the observed results, it was found that AF was efficient in elucidating the relationships between the characteristics and originating new variables. The factors formed are interesting in terms of selection, because in addition to allowing for joint interpretations, they present good estimates of predictive capacity, heritability and accuracy. Furthermore, there was a high agreement between the individuals selected based on the factors and those selected considering the individual variables. However, it is worth noting that, the selection based on the factors managed to select individuals of more appropriate size. Keywords: Genomic Prediction. Multivariate Analysis. Breeding Genetic.
Brazil stands out worldwide in the production of coffee. The increases observed in its productivity are the result of the improvement of several methodologies. Among them, the predictive methods of genetic value stand out. These contribute significantly to the selection of superior genotypes, in order to increase the genetic gain per unit of time. In this context, broad genomic selection (GWS) is a tool that stands out, since it allows predicting the future phenotype of an individual based only on information from molecular markers. Performing the selection simultaneously for various characteristics is the interest of most breeding programs, and factor analysis (AF) has been used to assist in this end. The use of factors is justified due to the existence of genetic correlations between the characteristics, which can be attributed to the QTL that have pleiotropic effects or to the closely linked QTL. Thus, the objective of this work was to evaluate the use of AF in the context of GWS, in genotypes of Coffea canephora. The results obtained from the selection based on the factors were compared, through predictive capacity, accuracy and the Cohen’s Kappa coefficient, with those derived from the analysis of the individual variables. For this, phenotypic and genotypic data from populations composed of clones of the varietal groups Conilon and Robusta and hybrids originated from crosses between these groups, evaluated for three consecutive years (2014 to 2016), and a density of 18111 identified SNPs markers were used. From the observed results, it was found that AF was efficient in elucidating the relationships between the characteristics and originating new variables. The factors formed are interesting in terms of selection, because in addition to allowing for joint interpretations, they present good estimates of predictive capacity, heritability and accuracy. Furthermore, there was a high agreement between the individuals selected based on the factors and those selected considering the individual variables. However, it is worth noting that, the selection based on the factors managed to select individuals of more appropriate size. Keywords: Genomic Prediction. Multivariate Analysis. Breeding Genetic.
Descrição
Dissertação de mestrado defendida na Universidade Federal de Viçosa.
Palavras-chave
Análise Multivariada, Predição, Genômica, Melhoramento Genético
Citação
PAIXÃO, Pedro Thiago Medeiros. Análise de fatores aplicada em estudos de seleção genômica no melhoramento de Coffea canephora. 2020. 35 f. Dissertação (Mestrado em Estatística Aplicada e Biometria) - Universidade Federal de Viçosa, Viçosa-MG. 2020.