Selection of maize (Zea mays L.) core colletions using an algorithm that maximizes allele richness and diverse genetic distance metrics

Abstract

Core collections (CC) are subsets of accessions that represent a broader collection, preserving their genetic diversity and avoiding redundancy. In plant breeding programs, CC are valuable tools since they enable the characterization and use of genetic resources without the need to evaluate the entire germplasm. In this context, the present work aimed to evaluate the implementation of a protocol to select multipurpose CC, with maximum genetic variability, to assist the INTA EEA-Pergamino Maize breeding program in future genomewide association mapping (GWAS) and genomic prediction (GP) studies. Core Hunter 3, an algorithm based on an advanced stochastic local search method, capable of incorporating genotypic and phenotypic matrices, was used to maximize various distance and genetic diversity parameters simultaneously. Seven CC of the same size (n=115) were selected with different optimization objectives based on genotypic data and its combination with phenotypic data, from a breeding panel (BP) consisting of 484 maize inbred lines. All selected CC retained 100% of the alleles present (CV=1), exhibiting similar average values for expected heterozygosity (HE=0.40) and observed heterozygosity (HO=0.02) as the BP. After evaluating the distribution of three phenotypic traits and the population representativeness of the CC selected by the algorithm, we determined that the CC selected with the A-NE objective meet the requirements to be used as multipurpose CC in future research investigations.