World Aquaculture Singapore 2026

June 2-5, 2026

Singapore

OPTIMISATION OF GENOMIC SELECTION FOR HARVEST TRAITS OF MALABAR RED SNAPPER Lujanus malabaricus

Bing Liang1,2, Dean R. Jerry2,4, Vu Nguyen2, Xueyan Shen2, Joyce Koh2,3, Celestine Terence2, Maria G. Nayfa2,3, Maura Carrai2, Kathiresan Purushothaman2,3,5, Rachel Ho Jia Wen2,3, Hazim Mohamed2,3, Saraphina Dianne Tneo Rwei Qing3, Grace Loo3, Shubha Vij2,3, Jose A. Domingos2

 

1Marine Aquaculture Centre, Singapore Food Agency, 52 Jurong Gateway Road, JEM Office Tower, #14-01, 608550, Singapore

2Tropical Futures Institute, James Cook University Singapore, 149 Sims Drive, 387380, Singapore

3School of Applied Science, Republic Polytechnic, 9 Woodlands Avenue 9, 738964, Singapore

4ARC Research Hub for Supercharging Tropical Aquaculture through Genetic Solutions, James Cook University, Townsville, QLD 4810, Australia

5Department of Preclinical Sciences and Pathology, Faculty of Veterinary Medicine, Norwegian University of Life Sciences, Ås, 1433, Norway

Email: Liang_Bing@sfa.gov.sg



Selective breeding in aquaculture has traditionally relied on pedigree-based methods, but these approaches are often less accurate and resource-intensive. Genomic selection (GS) offers a more precise alternative, using single nucleotide polymorphisms (SNPs) to estimate genomic estimated breeding values (GEBVs). This study optimized GS for Malabar red snapper (Lutjanus malabaricus), a commonly farmed tropical fish in Southeast Asia, by evaluating three parametric models: GBLUP (genomic best linear unbiased prediction), BayesR (a hierarchical Bayesian mixture model), and KAML (kinship-adjusted multiple-loci linear mixed model). These models were assessed across a range of SNP densities (500 to 56,378) and training population sizes (80 to 2,038). A total of 2,547 individuals from three rearing sites were genotyped using a 70k SNP array, assessing five traits: body weight (BW), total length (TL), body depth (BD), Fulton’s condition factor (K), and body shape index (BSI). Prediction accuracy was evaluated through five-fold cross-validation with 10 replicates.  Results showed similar performance across models for growth traits (BW, TL, and BD), with accuracies between 0.4 and 0.6. BayesR excelled in ratio traits, reaching 0.76 for K and 0.74 for BSI. Correlation analysis revealed strong alignment between GBLUP and KAML (≥0.95), while BayesR correlated highly with GBLUP and KAML for BW and BD (>0.93) but moderately for TL and BSI (0.54–0.74) and showed no significant correlation for K. Prediction accuracy improved with larger training populations and higher SNP densities, reaching plateau at 5,000 SNPs and 1,200 individuals for growth traits.  Additionally, top-ranked SNP subsets based on GWAS for BW outperformed random subsets in EBV prediction for BW, TL, and BD, but not for K and BSI. This suggests that targeted SNP selection is trait-specific and shows limited compatibility with unrelated traits.  Consequently, a balanced approach using 5,000 SNPs and a training population of 1,200 individuals was found to be the most cost-effective for GS predictions in Malabar red snapper.

In a complementary study , in-silico  pooled DNA genotyping was tested for progeny testing and genomic selection . Simulated mass spawning produced genotypes and phenotypes for parents and offspring.  Offspring with similar phenotypes were grouped into bands and genotyped by pool, and SNP allele frequencies were used to estimate parental contributions.  Assignment accuracy was above 99% with a minimum of 63 SNPs for 64 families. The approach  also  accurately estimated parental breeding values with 99% accuracy by genotyping only 10 pools instead of 1,200 individual offspring, reducing costs 120-fold.  This method can aid in genetic diversity analysis of the pool by assessing the proportional contributions of each sire and dam, offering valuable insights for offspring selection. In summary, this approach holds significant potential for reducing costs in the real-world genomic selection of both parents and offspring.