2021年南美白对虾体长和腹长/头胸长全基因组预测数据集

南美白对虾是世界范围内水产养殖业中最重要的品种之一，其中生长被视为选择性育种计划的首要性状。本研究分析了体长(BL)和腹长/头胸长(AL/CL)两种生长性状的遗传力和遗传相关性，并对基于机器学习等不同基因组选择模型的基因组预测进行了评价。BL和AL/CL的遗传力分别为0.25±0.04和0.07±0.03。两种表型呈中度负相关(- 0.70±0.14)。不同预测模型的比较表明，NeuralNet的预测精度最高。与GBLUP相比，NeuralNet的预测精度提高了约10%。在不同的标记密度下，NeuralNet的预测精度最高，使用1000个snp的预测精度与使用总snp的预测精度相近。在比较多性状模型(MTM)和单性状模型(STM)时，NeuralNet的预测精度提高了30%左右，优于其他方法。综上所述，NeuralNet模型在对虾基因组选择育种中具有较好的应用前景。这些结果为加快基因组选择育种在对虾改良计划中的应用提供了坚实的基础。

Litopenaeus vannamei is one of the most important aquaculture species worldwide, with its growth regarded as the primary trait in selective breeding programs. This study analyzed the heritability and genetic correlation of two growth traits, body length (BL) and abdominal length/carapace length (AL/CL), and evaluated genomic prediction performance across different genomic selection models including machine learning-based approaches. The heritability estimates for BL and AL/CL were 0.25 ± 0.04 and 0.07 ± 0.03, respectively. The two phenotypes exhibited a moderate negative correlation of -0.70 ± 0.14. Comparison of prediction models showed that NeuralNet achieved the highest prediction accuracy. Compared with GBLUP, the prediction accuracy of NeuralNet was improved by approximately 10%. Under varying marker densities, NeuralNet consistently maintained the highest prediction accuracy, and the accuracy using 1000 SNPs was comparable to that using the full set of SNPs. When comparing multi-trait model (MTM) and single-trait model (STM), the prediction accuracy of NeuralNet was increased by about 30%, outperforming all other methods. In summary, the NeuralNet model has promising application prospects in genomic selective breeding of Litopenaeus vannamei. These findings provide a solid foundation for accelerating the deployment of genomic selection breeding in shrimp improvement initiatives.