Deep learning methods such as multilayer perceptrons (MLPs) and convolutional neural networks (CNNs) have been applied to predict the complex traits in animal and plant breeding. However, it remains challenging to improve genomic prediction accuracy. To that end, a team of researchers from China applied CNNs to predict swine traits using previously published data.
Their results are reported in the Journal of Integrative Agriculture.
"We evaluated the CNN model's performance by employing various sets of single nucleotide polymorphisms (SNPs) and concluded that the CNN model achieved optimal performance when utilizing SNP sets comprising 1,000 SNPs," shares corresponding author Zhonglin Tang, a professor at Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences. "Furthermore, we adopted a novel approach using the one-hot encoding method that transforms the 16 different genotypes into sets of eight binary variables."
The team's innovative encoding method markedly enhanced the CNN's prediction accuracy for swine traits, outperforming the traditional one-hot encoding techniques. "Our findings suggest that the expanded one-hot encoding method can improve the accuracy of deep learning methods in the genomic prediction of swine agricultural economic traits," adds Tang. "These have significant implications for swine breeding programs, where genomic prediction is pivotal in improving breeding strategies."
The researchers recommend future research endeavors explore additional enhancements to deep learning methods by incorporating advanced data pre-processing techniques.
