GENOMIC PREDICTION AND TRAINING SET OPTIMIZATION AS ALTERNATIVES TO ACCELERATE THE BREEDING OF Acrocomia aculeata.

Acrocomia aculeata is a neotropical oil palm with great potential for biodiesel production. The species has gained prominence as sustainable agriculture becomes crucial for mitigating climate change. However, there is no improved germplasm since this palm has not been domesticated. Thus, to enhance the breeding of A. aculeata, we utilized genomic prediction models and applied training set optimization to select individuals from a subset that more accurately captures the diversity within a population. Three SNP calling methods were compared, and for the first time, we present the predictive accuracies of three productivity-related traits: fruit dry mass (FDM), pulp dry mass (PDM), and pulp oil content (OC). Data were collected from 201 trees over two years of evaluation. The trees were genotyped based on three references: the oil palm reference genome, de novo sequencing, and the A. aculeata transcriptome. The predictive ability of the GBLUP and BayesB models in cross-validation and real validation procedures was evaluated. Subsequently, various optimization criteria were tested to verify consistency and the ability to provide an optimized training set that reduced risk in both targeted and untargeted scenarios. Using the oil palm genome as a reference, together with GBLUP models, yielded the best results for genomic prediction of FDM, OC, and PDM, with accuracies of 0.46, 0.45, and 0.39, respectively. The PEV, r-score, and core collection methodology criteria provided more secure decisions. Optimizing the training set serves as an effective strategy to enhance decision-making in the pre-breeding process of A. aculeata, leveraging genomic data as a cost-efficient means to accelerate its domestication. The predictive performance of genomic models tends to remain consistent when utilizing an optimized training set, offering breeders the confidence needed to use them as a reference for characterizing populations of A. aculeata.