Trends in the Application of Multiomics Based on Machine Learning in the Development of Probiotics

With the rapid development of computational methods and high-throughput multiomics technologies, machine learning (ML) has emerged as an important analytical approach in probiotic research. This review summarizes recent ML-assisted applications across genomics, transcriptomics, metabolomics, microbiome profiling, and culturomics, and organizes current studies around four functional objectives: probiotic selection, functional prediction, metabolic activity prediction, and probiotic effectiveness optimization. We discuss how ML facilitates the integration of heterogeneous omics data to enable more systematic and quantitative probiotic development and highlight representative analytical tools and workflows. At the same time, key limitations remain, including cross-platform data heterogeneity, imbalanced functional labels, and limited robustness in capturing complex microbial and environmental interactions. Consequently, experimental validation remains essential for ensuring biological relevance. Future progress will rely on standardized multiomics integration and iterative computational-experimental frameworks to support rational probiotic optimization.