Combining genotyping-by-sequencing data sets for GWAS and GWS study for maize root System architecture traits

• Ano de defesa: 2019
• Autor(a) principal: Zuffo, Leandro Tonello
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Zea mays; Milho - Seleção; Genômica; Miho - Mapeamento genômico; Milho - Populações; Melhoramento Vegetal
• Link de acesso: https://locus.ufv.br//handle/123456789/28822

Maize has a sophisticated and complex root architecture that is important for plant anchorage and uptake of nutrients and water. Tools such as genome-wide association (GWAS) and genome-wide selection (GWS) can help the breeders to make decisions about which genotypes should advance in their breeding programs. In this work, we combined two data sets with different genetic backgrounds to perform a GWAS and GWS analysis in different scenarios. Our hypothesis was that combining the genomic data from different backgrounds may increase the power of detecting significant polymorphisms associated with maize seedling root traits and increase prediction accuracy. We used 679 inbred lines, 377 from Ames Panel and 302 from BGEM Panel. The root seedlings phenotype was obtained via image software from plants 14 days old grown in paper rolls in a growth chamber. We evaluated five root traits: depth, lateral root length, primary root length, total number of roots and total root length. Our study is singular by combining two SNP (Single Nucleotide Polymorphism) data sets with different genetic backgrounds to access the prediction accuracy within, across and combining the populations or subpopulations on root traits in maize. After quality control, 232,460 SNPs were used in further analysis. Population structure analysis revealed four groups that were used to build the scenarios for GWS and to control false positives in GWAS analysis. GWAS showed a total of 13 SNPs above the significance threshold. Those SNPs led to 10 candidate genes. At GWS, the combined scenario had the highest accuracy (0.66) across all traits, followed by non- stiff-stalk combined (0.63) and stiff-stalk combined (0.56). The scenarios that we calculated the prediction accuracy across panels showed the low accuracies (all were lower than 0.25). As seen in this study, combine results across studies is useful, even with different backgrounds, to improve the GWS accuracy and detect significant polymorphisms associated with maize seedling root traits and allocate the individuals from the combined data set in groups by using population structure analysis is advantageous. The genes found can be further studied to help understand the genetic basis of root development and improve the root architecture. Keywords: Zea mays. Genomic selection. Association Mapping. Population structure