New strategies for implementing of genomic selection in breeding programs of clonally propagated crops

• Ano de defesa: 2019
• Autor(a) principal: Batista, Lorena Guimarães
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Allele dosage; Cana-de-açúcar; Dosagem alélica; Genomic selection; Índice de seleção; Poliploides; Polyploids; Seleção genômica; Selection index; Sugarcane
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/11/11137/tde-30072019-111124/

Genomic selection consists of using predicted effects of genetic markers to predict breeding values and/or genotypic values of genotyped individuals. With this approach, selection can be carried based only on those predicted breeding values, reducing the need for further phenotypic evaluations. This represents a great advance in terms of cost and effectiveness of selection in breeding programs of all kinds of crops. In the first chapter of this work, we explore one of the ways genomic selection can be used to increase efficiency when breeding clonally propagated crops for multiple traits. Using stochastic simulations, we show that an economic selection index should be preferred over independent culling. Our results show that the use of genomic selection may render the cost-efficiency benefit of independent culling obsolete when all early generation individuals are genotyped and accurate prediction of all traits becomes available simultaneously. Despite the potential benefits of selecting based on predicted breeding values, for some clonally propagated species the complexity of their genomes limits the implementation of genomic selection in breeding programs. Since including allele dosage information has been shown to improve performance of genomic selection models in autotetraploid species, our objective in the second chapter of this work was to assess the accuracy of genome-wide prediction in the highly complex polyploid sugarcane when incorporating allele dosage information. In this chapter, we expanded GBLUP genomic selection models developed for autotetraploids to include higher levels of ploidy. Two types of model were used, one with additive effects only and one with additive and digenic dominance effects. We observed a modest improvement in the performance of the prediction model when ploidy and allele dosage estimates were included, indicating that this is a possible way of improving genomic selection in sugarcane. The results obtained in both studies can assist researchers and breeders of clonally propagated crops, opening new research opportunities and indicating the most efficient ways to implement genomic selection.