Bayesian genomic models boost prediction accuracy for survival to Streptococcus agalactiae infection in Nile tilapia (Oreochromus nilioticus)

Abstract Background Streptococcosis is a major bacterial disease in Nile tilapia that caused by Streptococcus agalactiae infection, and development of resistant strains represents sustainable approach towards combating this disease. In study, we performed controlled trial on 120 full-sib families to (i) quantify characterize the potential genomic selection for survival S. infection tilapia, (ii) identify best model optimal density single nucleotide polymorphisms (SNPs) trait. Methods total, 40 fish per family (15 intraperitoneally injected 25 as cohabitants) were used challenge test. Mortalities recorded every 3 h 35 days. After quality control, genotypes (50,690 SNPs) phenotypes (0 dead 1 alive) 2472 cohabitant available. Genetic parameters obtained using various models (genomic linear unbiased prediction (GBLUP), BayesB, BayesC, BayesR BayesS) traditional pedigree-based (PBLUP). The analysis deep 17-generation pedigree. Prediction accuracy bias evaluated five replicates tenfold cross-validation. further analyzed 10 subsets SNPs at different densities explore effect pruning SNP predictive accuracy. Results Moderate estimates heritabilities ranging from 0.15 ± 0.03 0.26 0.05 with models. Compared model, GBLUP (using all increased 15.4%. Furthermore, use most appropriate Bayesian up 71%. 50 non-zero effects consistent BayesC BayesS respect marker id and/or locations. Conclusions These results demonstrate tilapia. PBLUP models, found boost significantly.