Computational Molecular Biology 2025, Vol.15, No.5, 235-244 http://bioscipublisher.com/index.php/cmb 238 as random forests and lift algorithms, often outperform old-school models in predicting milk production phenotypes. Although neural networks and deep learning are complex, they perform well in high-dimensional data analysis, especially in feature selection, and can screen out which SNPS, genes or metabolites have the greatest impact on traits (Frizzarin et al., 2021). Of course, it's not the case that the more complex the model is, the better. The key is to ensure that the results are stable. Therefore, if multi-omics data can be reasonably combined with AI methods, it is not difficult to make stable phenotypic predictions and precise breeding decisions. Figure 1 The proportion of variance explained by lactation associated RNA editing sites (Adopted from Cai et al., 2025) 5 Identification of Key Pathways and Regulatory Networks 5.1 Signaling pathways in mammary gland development and lactation regulation During the process of breast development and lactation, more than one signaling pathway is at play. Pathways such as PI3K-Akt, AMPK, insulin and PPAR are often simultaneously involved in cell proliferation, fat synthesis
RkJQdWJsaXNoZXIy MjQ4ODYzNA==