Análise genética de características de produção in vitro de embriões de doadoras da raça Gir Leiteiro
Carregando...
Arquivos
Data
2019-11-08
Autores
Pereira, Mariana Alencar
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O presente estudo teve como objetivo: 1) estimar parâmetros genéticos para características de produção in vitro de embriões de doadoras da raça Gir Leiteiro considerando as distribuições Gaussiana e Poisson; 2) investigar a existência de regiões cromossômicas que possam estar associadas com essas características. Os dados utilizados totalizaram 12.491 informações de aspirações foliculares provenientes de 5.496 doadoras da raça Gir Leiteiro. As características número de oócitos viáveis (NOV) e número de embriões viáveis (NEV) foram analisadas por meio de modelo animal unicaracterísticas considerando os efeitos sistemáticos de grupo de contemporâneos (data da aspiração folicular e rebanho), idade da doadora (covariável linear e quadrática) e intervalo entre as aspirações (OPU) (covariável linear). Os efeitos de animal e de ambiente permanente, além do residual, foram assumidos como aleatórios. Os componentes de variância foram estimados via modelos lineares generalizados mistos sob o enfoque bayesiano. A herdabilidade estimada para NOV foi de 0,12 para os modelos Gaussiano e Poisson, respectivamente e para NEV, as estimativas de herdabilidade foram de 0,02 e 0,03, respectivamente. Os valores de repetibilidade variaram em função do modelo estudado, de 0,17 a 0,22 para NOV e de 0,08 a 0,37 para NEV, empregando-se os modelos Gaussiano e Poisson, respectivamente. O modelo Poisson ajustou-se melhor para as características estudadas, uma vez que apresentou menores valores para o critério de informação da deviance (DIC). Os valores genéticos preditos para as características obtidos pelos diferentes modelos foram altamente correlacionados, 0,79 para NOV e 0,87 para NEV. Considerando os 10% melhores animais de acordo com o valor genético, observou-se um maior re-ranqueamento para a característica NOV em relação à NEV, indicando que a escolha do modelo adequado é importante para o processo de seleção dos animais. A partir dos valores genéticos preditos via modelo Poisson foram calculados, de forma iterativa, os efeitos dos marcadores ponderados. Os resultados da análise de GWAS via GBLUP foram apresentados com base no percentual de variância explicada por janelas de tamanho de 100 SNPs adjacentes. Apenas SNPs com variância genética aditiva acima de 1% foram considerados. A variância explicada pelas janelas de maior efeito foi de 21,22% e estas estavam localizadas nos cromossomos 1, 2, 3, 4, 5, 8, 13, 14, 17, e 27 para a característica NOV. Para a característica NEV, a variância explicada pelas janelas de maior efeito foi de 13,39% e estas estavam localizadas nos cromossomos 2, 6, 9, 14, 18 e 25. Nessas regiões foram identificados genes relacionados direta ou indiretamente com as características estudadas.
The present study aimed to: 1) estimate genetic parameters for in vitro production traits of Dairy Gyr donors considering Gaussian and Poisson distribution; 2) investigate the existence of chromosomal regions that may be associated with these traits. The data totalized 12,491 records of ovum pick-up from 5,496 Dairy Gyr donors. The traits viable oocyte number (VON) and viable embryos number (VEN) were analyzed by unitrait animal model considering systematic effects of contemporary group (ovum pick-up date and herd), age of donor (covariable linear and quadratic) and pick-up interval (OPU) (covariable linear). Additive genetic, permanent environmental and residual effects were assumed as random. The variance components were estimated by generalized linear mixed models under Bayesian approach. The heritability estimated for VON was 0.12 for both models. For VEN estimated heritability was 0.02 and 0.03 for Gaussian and Poisson models, respectively. The repeatability ranged according to the model studied, 0.17 to 0.22 for VON and 0.08 to 0.37 for VEN, employing Gaussian and Poisson models, respectively. Poisson model showed the best fit to the studied traits, according the deviance information criterion (DIC). The predicted breeding values of traits using the different models were high correlated, 0.79 to VON and 0.87 to VEN. Considering the 10% best breeding values the re-ranking of animals for VON was higher when compared to VEN, suggesting that choosing an appropriated model is important for selection process of animals. The results of GWAS analysis through GBLUP were based on the percentage of variance explained by windows of 100 adjacent SNPs. Only SNPs with additive genetic variance above 1% were considered. The variance explained by the window with larger effects was 21.22%, observed on chromosomes 1, 2, 3, 4, 5, 8, 13, 14, 17 and 27 for NOV. For the NEV trait, the variance explained by the windows with larger effects was 13.39%, on chromosomes 2, 6, 9, 14, 18 and 25. In these regions genes directly or indirectly related to the studied traits were identified.
The present study aimed to: 1) estimate genetic parameters for in vitro production traits of Dairy Gyr donors considering Gaussian and Poisson distribution; 2) investigate the existence of chromosomal regions that may be associated with these traits. The data totalized 12,491 records of ovum pick-up from 5,496 Dairy Gyr donors. The traits viable oocyte number (VON) and viable embryos number (VEN) were analyzed by unitrait animal model considering systematic effects of contemporary group (ovum pick-up date and herd), age of donor (covariable linear and quadratic) and pick-up interval (OPU) (covariable linear). Additive genetic, permanent environmental and residual effects were assumed as random. The variance components were estimated by generalized linear mixed models under Bayesian approach. The heritability estimated for VON was 0.12 for both models. For VEN estimated heritability was 0.02 and 0.03 for Gaussian and Poisson models, respectively. The repeatability ranged according to the model studied, 0.17 to 0.22 for VON and 0.08 to 0.37 for VEN, employing Gaussian and Poisson models, respectively. Poisson model showed the best fit to the studied traits, according the deviance information criterion (DIC). The predicted breeding values of traits using the different models were high correlated, 0.79 to VON and 0.87 to VEN. Considering the 10% best breeding values the re-ranking of animals for VON was higher when compared to VEN, suggesting that choosing an appropriated model is important for selection process of animals. The results of GWAS analysis through GBLUP were based on the percentage of variance explained by windows of 100 adjacent SNPs. Only SNPs with additive genetic variance above 1% were considered. The variance explained by the window with larger effects was 21.22%, observed on chromosomes 1, 2, 3, 4, 5, 8, 13, 14, 17 and 27 for NOV. For the NEV trait, the variance explained by the windows with larger effects was 13.39%, on chromosomes 2, 6, 9, 14, 18 and 25. In these regions genes directly or indirectly related to the studied traits were identified.
Descrição
Palavras-chave
Fertilização in vitro, Herdabilidade, Inferência bayesiana, Modelos lineares generalizados mistos, Polimorfismos de um único nucleotídeo, Zebu