Caracterização e associação genética de linhagens e predição genômica da produtividade de grãos de híbridos de milho utilizando marcadores DARTS-GBS

Abstract

Corn culture is extremely important for human and animal feeding in the entire world. Thus, its breeding, aiming at obtaining more productive hybrids, is essential when facing increasing demands for food. For this, it is necessary to understand in depth the strain collection used for breeding and the best combinations between them. Therefore, in the first part of this work, we aimed at exploring the information obtained from genotyping a germplasm bank derived from a seed company, aiming at its detailing; at evaluating the genetic diversity and population structure of the collection; and investigate the potential of the collection as source for studying the genetic architecture of corn ear weight. In the second part of this work, the objective was the genome wide selection using a set of Darts-seq markers and the GLUP model with dominance for grain productivity of corn hybrids evaluated in different seasons and locations. We conducted DNA extraction and posterior genotyping of accesses from the collection. We performed genetic grouping of the inbred lines of the collection and genomic association analysis of corn ear weight. From the crosses of the inbred lines, 838 hybrids were obtained, evaluated at six locations in the winter season of 2013, and 797 hybrids, evaluated at four locations in the summer season of 2013/2014. We also used the common hybrids from both seasons to calculate the correlations between them. We verified that the germplasm bank presented no fixed allele frequencies, demonstrating the selection and breeding potential of the genotypes. It was possible to separate the inbred lines into distinct heterotic groups, with the majority consistent with the pedigree information. However, the set of data used for the association study was inadequate in identifying significant effect marks for corn ear weight. Coincidences occurred between the VGGs of the summer and winter seasons. The GBLUP method was capable of generating high correlations between predicted and observed hybrids, even with high levels of unbalance and different locations and years.