First results of genome-wide association study for the fatty acids composition in olive oil

The cultivated olive tree is the oldest and most widespread tree crop in the Mediterranean basin. The focus of most of the olive production is to make olive oil, which is popular due to its chemical composition. Olive oils with a high content of oleic acid and a low content of linoleic acid are more desirable. Consequently, developing new cultivars that produce oils with a high oleic/linoleic acid ratio is a priority in olive breeding programs. Understanding the basis of quantitative traits can help plant breeders improve yields, fortify tolerance to stressors, and maximize the expression of qualitative traits. In recent years, association mapping (AM) methods have been developed to link individual genetic backgrounds (i.e., genotypes) with specific traits (i.e., phenotypes) based on linkage disequilibrium. In this work, a panel of 142 cultivars were genotyped using the single primer enrichment technology (SPET) and the fatty acid composition of olive oil was determined by nuclear magnetic resonance (NMR). Since variants calling was based on the assembly of the 'Leccino' genome, this was kept as a reference through the analysis with respective gene models. The raw VCF file, comprising 4,532,192 variants, was filtered, resulting in 142 individuals with 479,208 variants. Of these, 6,949 were in coding sequences of 1,055 genes belonging mainly to the biosynthetic pathways of primary metabolism. Multidimensional scaling (MDS) was used to graphically show distances between individuals in the population using each individual's coordinates along the principal axes of a MDS graph. These coordinates were used as control variables (i.e., covariates) in genotype/phenotype association tests. LD pruning returned a set of high-quality variants (n=131,554) that were used as input by ADMIXTURE for the estimation of allele frequencies and classification of individuals into six groups. In the genotype-phenotype association tests, five statistical models were used and 8 SNPs were found to be significantly associated with fatty acid composition. These first results may help to clarify the genetic basis explaining the variability of the fatty acid composition in olive trees and provide new molecular tools for molecular-assisted selection.