Related Experiment Video
Updated: May 25, 2026
Using RNA-mediated Interference Feeding Strategy to Screen for Genes Involved in Body Size Regulation in the Nematode C. elegans
Published on: February 13, 2013
Genetic parameters estimation and optimization of genomic selection in mud crab (Scylla paramamosain): a case study
Xiyi Zhou1,2,3, Zhenying Hong1,2,3, Wenxiao Cui1,2,3
1Guangdong Provincial Key Laboratory of Marine Biotechnology, Shantou University, 243 Daxue Road, Shantou, 515063, China.
Background:
Genomic selection (GS) has become a preferred method for genetic improvement in aquaculture species. However, its application in mud crab (Scylla paramamosain) remains limited. Given that the prediction accuracy of GS varies across species and is influenced by several factors, including statistical models, marker density, and the size of the reference population, a systematic investigation of GS in mud crab is necessary. To explore the impact of these factors on GS and optimize the application strategies of GS in mud crab breeding, a total of 508 mud crabs were genotyped by the 40 K liquid SNP array (named "Xiexin No. 1") to estimate genetic parameters of growth traits and systematically investigate the impact of prediction models, SNP density, and reference population size on GS.
Results:
Results showed that the genomic heritability for body weight (BW), carapace length (CL), carapace width (CW), and body height (BH) was 0.860 ± 0.043, 0.554 ± 0.081, 0.521 ± 0.083, and 0.597 ± 0.077, respectively. Different models (GBLUP, rrBLUP, BayesA, BayesB, BayesC, and BayesR) exhibited similar prediction accuracies for four growth-related traits, with ranges of 0.510-0.515, 0.569-0.574, 0.567-0.570, and 0.543-0.548 for BW, CL, CW, and BH, respectively, but GBLUP showed advantages in computational efficiency. Notably, the accuracy of GS for four growth-related traits would be further improved as the SNP density increased but began to plateau after 10 K SNPs. When SNP density increased from 0.5 K to 33 K, the average prediction accuracies of six models for BW, CL, CW, and BH were improved by 6.22%, 4.20%, 4.40%, and 5.23%, respectively. Moreover, the prediction accuracy of GS was also improved as the number of reference individuals increased. When the reference population size expanded from 30 to 400, the average prediction accuracies of six models for BW, CL, CW, and BH increased by 8.66%, 3.99%, 4.97% and 4.56%, respectively. In addtion, the prediction unbiasedness of GBLUP, BayesA, BayesB, and BayesC were close to 1, only when the reference population size was more than 150.
Conclusions:
Taken together, the reference population comprising at least 150 samples genotyped with over 10 K SNPs is the minimum standard for implementing GS of growth-related traits in mud crab. Our results would provide valuable insights and practical recommendations on implementing GS in mud crab genetic breeding programs.
