Animal Molecular Breeding, 2025, Vol.15, No.2, 49-59 http://animalscipublisher.com/index.php/amb 53 show that compared with the traditional BLUP model, the accuracy of predicting yield-related traits with GBLUP and ssGBLUP has increased by 10% to 13%. Bhat et al. (2023) and Negro et al. (2024) hold that although AI models are better at handling complex gene interactions and have the potential to further enhance predictive performance, these models have high requirements for computing power and still need to undergo rigorous verification before truly replacing traditional methods. Figure 2 Potential of artificial intelligence (AI)-based machine-learning (ML) and deep-learning (DL) models in genome-wide associationstudies (GWAS) and genomic selection (GS) analyses. The AI-based models capture linear and nonlinear interactions in GWAS and GS for use incrop breeding; MTAs represent the marker-trait associations (Adopted from Bhat et al., 2023) 6 Framework for Integrating AI with Gene Editing 6.1 AI for target site identification and efficiency prediction AI has been used in plant breeding to predict what impact a certain genetic variation will have on external performance, helping scientists better design gene editing. Farooq et al. (2024) hold that this method is actually also applicable to many animals such as goats. AI is very strong in pattern recognition and big data analysis. Researchers can use it to predict more accurately which parts will be edited correctly and also discover potential "off-target" problems. AI can also help optimize the design of gRNA in CRISPR/Cas9, improve editing efficiency and reduce the occurrence of unexpected mutations (Zhang et al., 2018; Zhang et al., 2019). 6.2 Functional genomics and AI With the development of high-throughput sequencing technology and SNP chips, scientists have obtained a large amount of genetic data. Subsequently, AI can conduct in-depth analysis of these data to help identify which genetic variations may be related to important traits such as disease resistance and reproductive ability (Rupp et al., 2016; Farooq et al., 2024). AI algorithms can also narrow the gap between genotypes and phenotypes, and identify candidate genes and regulatory elements that may affect traits. 6.3 Designing gene editing strategies with AI assistance Artificial intelligence (AI) can integrate genomic, expression and functional data to help design and optimize gene editing programs. AI can prioritize the selection of appropriate editing targets based on the influence of target traits and safety requirements, which is of great value for breeding more disease-resistant or high-yielding goats (Li et al., 2024). AI models can also simulate different gene editing schemes, helping researchers select the most
RkJQdWJsaXNoZXIy MjQ4ODYzNA==