Uso de haplótipos e SNPs em estudos de seleção e associação genômica para características reprodutivas em bovinos da raça Nelore
Carregando...
Data
2019-04-30
Autores
Pinzón, Andrés Chaparro
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
As características reprodutivas são fundamentais para a rentabilidade do sistema produtivo de gado de corte. Contudo, características como idade ao primeiro parto (IPP) e perímetro escrotal (PE) possuem desvantagens para serem utilizadas em programas de melhoramento genético tradicional, uma vez que são mensuradas em só um sexo e podem apresentar baixa herdabilidade. Nas últimas décadas os avanços nas tecnologias de análise de DNA têm permitido o desenvolvimento de procedimentos estatísticos como estudos de associação (GWAS) e seleção genômica (GS). Comumente têm-se utilizado marcadores de tipo SNP para desenvolver este tipo de estudos. Entretanto, outro tipo de marcador molecular os haplótipos, que são grupos de SNPs que estão em alto desequilíbrio de ligação (DL) podem ser utilizados neste tipo de estudo, uma vez que estes podem estar em maior desequilíbrio de ligação com os QTL quando comparados com os SNPs. O objetivo deste estudo foi verificar o uso de haplótipos em estudos de GWAS e GS, para características reprodutivas, em bovinos da raça Nelore. Para o estudo de GWAS, 2.390 observações de IPP e 4.832 informações para CE, provenientes de três programas de melhoramento da raça Nelore, foram utilizadas em um estudo de associação genômica ampla. 1900 fêmeas e 1500 machos jovens foram genotipados com o painel HD da Ilumina® (777K). 490 fêmeas e 3.332 machos jovens foram genotipados utilizando o painel da GeneSeek 75K. Os animais genotipados com painel de menor densidade foram imputados para HD utilizando o programa FImpute. O fenótipo foi ajustado para os efeitos fixos (Y*). Os haplótipos foram construídos por médio do software HaploView utilizando desequilíbrio de ligação (D') e a metodologia de intervalo de confiança. Para o GWAS foi utilizado um modelo misto univariado no software GEMMA. Vinte e um blocos significativos de haplótipos ao longo de 12 diferentes cromossomos bovinos foram observados. Dez blocos localizados nos cromossomos 5, 9, 11, 13, 19, 21, 22 e 25 foram associados a genes que têm efeitos putativos nos sistemas reprodutivos em mamíferos. Os genes FAF1, HDAC4, ARHGEF13, CCDC62, SLC25A39, FKBP6 e NSUN5 foram associados com CE e estão envolvidos em funções reprodutivas como espermatogênese, maturação e mobilidade dos espermatozoides. Os resultados encontrados no presente estudo podem ser uteis para identificar mutações causais e genes candidatos que influenciam as características IPP e CE. No estudo de seleção genômica, fenótipos e genótipos de 2.390 e 4.832 para IPP (h2 =0,08) e circunferência escrotal (CE, h2 =0,42), respectivamente, foram analisados. A metodologia para construir os haplótipos foi a mesma descrita no capítulo 2. Foi realizada validação cruzada com 5 grupos e o método GBLUP foi usado para predição do valor genômico usando haplótipos e SNPs como preditores. A acurácia e o viés foram empregados para comparar a habilidade de predição dos haplótipos e SNPs. A acurácia de predição para IPP foi de 0,16 para SNPs e haplótipos e para CE foi de 0,21 para SNPs e 0,22 para haplótipos. O coeficiente de regressão para IPP foi < 1 e para CE foi > 1, indicando que a predição genômica pode estar subestimada para IPP e superestimada para CE. Valores similares de coeficientes de regressão foram encontrados para ambas as características quando comparados haplótipos e SNPs. O coeficiente de herdabilidade afetou as acurácias de predição, causando um pequeno aumento usando haplótipos em CE (0.42) e igual acurácia entre os preditores para IPP (0.08)
The reproductive traits are fundamental for the profitability of the beef cattle production system profitability. However, traits such as age at firs calving (AFC) and scrotal circumference (SC) have disadvantages to be used in traditional breeding programs, since they are measured in only one sex and may have low heritability. In the last few decades, the advances in technology for DNA analysis allowed the development of statistical techniques as genomic-wide (GWAS) and genomic selection studies (GS). Commonly, SNPs markers have been used to perform these studies. Nonetheless, another molecular marker, the haplotype, that are SNPs in high linkage disequilibrium (LD), could be used in these studies, since these could be in higher LD with the QTLs when compared to the individual SNPs. The aim of this study was to verify the use of haplotypes in GWAS and GS, for reproductive traits, in Nelore cattle. In GWAS study, 2,390 and 4,832 animals with information of age at first calving (AFC) and scrotal circumference (SC) belonging to three Nelore breeding programs were used to perform the GWAS. The genotypes of 1900 heifers and 1500 young bulls were obtained with HD panel from Ilumina® (777K) and 490 heifers and 3332 young bulls were genotyped with the GeneSeek Genomic Profiler HDi 75K. Animals genotyped with lower density panel were imputed to HD using the FImpute program. Phenotype was adjusted for the contemporary groups fixed effects (Y*). Missing genotypes and linkage phase were inferred with the Fimpute software. Haplotypes were constructed in the software HaploView using the linkage disequilibrium measurement D’ and the confidence interval methodology. The GWAS was performed using univariate mixed model in the software GEMMA. Twenty-one significant haplotype blocks along 12 different chromosomes were associated with AFC. Ten blocks located in the chromosomes 5, 9, 11, 13, 19, 21, 22, and 25 were associated with genes that have putative effects on reproductive systems in mammals. The genes FAF1, HDAC4, ARHGEF13, CCDC62, SLC25A39, FKBP6 and NSUN5 were associated with SC and are involved in reproductive functions as spermatogenesis and spermatozoa maturation and motility. The results found in this study can be used to identify causal mutations and candidate genes that underline AFC and SC. In chapter 3, phenotypes and genotypes of 2,390 and 4,832 animals were available for AFC (h²=0.08) and scrotal circumference (SC, h²=0.42) respectively. The methodology used to impute the missing genotypes, the haplotype phase and construction was descripting in chapter 2. A cross-validation with 5 folds was performed and GBLUP method was used for genomic prediction using haplotypes and SNPs as predictors. The accuracy and bias were used to compare the prediction models. The accuracy of prediction for AFC was 0.16 for SNPs and haplotypes and for SC was 0.21 for SNPs and 0.22 for haplotypes. The regression coefficient for AFC was >1 and for SC was <1, showing that the prediction accuracy was overestimating for AFC and underestimating for SC. Similar regression coefficient values were found for both traits when compared haplotypes and SNPs. The heritability coefficient affected the prediction accuracies, causing slight increase using haplotypes for trait of high heritability (0.42) and equal accuracy between predictors for trait of low heritability (0.08).
The reproductive traits are fundamental for the profitability of the beef cattle production system profitability. However, traits such as age at firs calving (AFC) and scrotal circumference (SC) have disadvantages to be used in traditional breeding programs, since they are measured in only one sex and may have low heritability. In the last few decades, the advances in technology for DNA analysis allowed the development of statistical techniques as genomic-wide (GWAS) and genomic selection studies (GS). Commonly, SNPs markers have been used to perform these studies. Nonetheless, another molecular marker, the haplotype, that are SNPs in high linkage disequilibrium (LD), could be used in these studies, since these could be in higher LD with the QTLs when compared to the individual SNPs. The aim of this study was to verify the use of haplotypes in GWAS and GS, for reproductive traits, in Nelore cattle. In GWAS study, 2,390 and 4,832 animals with information of age at first calving (AFC) and scrotal circumference (SC) belonging to three Nelore breeding programs were used to perform the GWAS. The genotypes of 1900 heifers and 1500 young bulls were obtained with HD panel from Ilumina® (777K) and 490 heifers and 3332 young bulls were genotyped with the GeneSeek Genomic Profiler HDi 75K. Animals genotyped with lower density panel were imputed to HD using the FImpute program. Phenotype was adjusted for the contemporary groups fixed effects (Y*). Missing genotypes and linkage phase were inferred with the Fimpute software. Haplotypes were constructed in the software HaploView using the linkage disequilibrium measurement D’ and the confidence interval methodology. The GWAS was performed using univariate mixed model in the software GEMMA. Twenty-one significant haplotype blocks along 12 different chromosomes were associated with AFC. Ten blocks located in the chromosomes 5, 9, 11, 13, 19, 21, 22, and 25 were associated with genes that have putative effects on reproductive systems in mammals. The genes FAF1, HDAC4, ARHGEF13, CCDC62, SLC25A39, FKBP6 and NSUN5 were associated with SC and are involved in reproductive functions as spermatogenesis and spermatozoa maturation and motility. The results found in this study can be used to identify causal mutations and candidate genes that underline AFC and SC. In chapter 3, phenotypes and genotypes of 2,390 and 4,832 animals were available for AFC (h²=0.08) and scrotal circumference (SC, h²=0.42) respectively. The methodology used to impute the missing genotypes, the haplotype phase and construction was descripting in chapter 2. A cross-validation with 5 folds was performed and GBLUP method was used for genomic prediction using haplotypes and SNPs as predictors. The accuracy and bias were used to compare the prediction models. The accuracy of prediction for AFC was 0.16 for SNPs and haplotypes and for SC was 0.21 for SNPs and 0.22 for haplotypes. The regression coefficient for AFC was >1 and for SC was <1, showing that the prediction accuracy was overestimating for AFC and underestimating for SC. Similar regression coefficient values were found for both traits when compared haplotypes and SNPs. The heritability coefficient affected the prediction accuracies, causing slight increase using haplotypes for trait of high heritability (0.42) and equal accuracy between predictors for trait of low heritability (0.08).
Descrição
Palavras-chave
Bos indicus, Circunferência escrotal, Gado de corte, Idade ao primeiro parto, Valores genéticos genômicos, Age at first calving, Beef cattle, Genomic breeding values, Scrotal circumference