Francisella orientalis infection causes major economic losses, making genetic improvement for resistance a necessity. Genomic selection (GS) can accelerate this process, but genotyping cost is the main limitation, especially in emerging countries. This study, therefore, evaluated strategies to create low densities SNP panels by prioritizing relevant markers, aiming to lower costs without sacrificing prediction accuracy.
A total of 920 Nile tilapia were challenged with Francisella orientalis, and resistance was measured as time to death. Of these, 461 were genotyped using a commercial 60K SNP panel. The 132 parents were low-pass sequenced (LP-Seq) and used as a reference panel to impute the challenged fish genotypes, generating a high-density panel of ~800K SNPs imputed. Different SNPs densities panels were created using Linkage Disequilibrium (LD) selection (20K, 1K, 200 SNPs) or by prioritizing SNPs from GWAS (46K, 20K, 1K, 200 SNPs). Predictive ability was assessed using pBLUP and ssGBLUP with 5-fold cross-validation.
This approach can significantly reduce genotyping costs, facilitating the widespread adoption of genomic selection to improve disease resistance and support a more sustainable aquaculture industry.