Multi-omics network analysis has become essential for understanding the complex and interdependent biological processes that drive human health and disease. As high-throughput technologies begin to characterize transcriptomic, epigenomic, proteomic, and metabolomic layers in the same subjects, the need for rigorous integrative methods has grown substantially. Without tools that capture coordinated and higher-order cross-omics relationships, critical regulatory mechanisms remain undetected, limiting both mechanistic insight and translational progress.

This Methods Collection aims to showcase innovative and adaptable methodologies for multi-omics data integration, particularly approaches that move beyond simple pairwise associations to model complex multi-view structure. By featuring tensor-based models, sparse and structured canonical correlation frameworks, autoencoder-based dimensional reduction, and clustering algorithms tailored for heterogeneous datasets, the Collection will highlight analytical strategies that address real-world challenges such as high dimensionality, nonlinearity, missingness, and diverse study designs. These methods will help researchers uncover shared biological embeddings, identify interpretable network modules, and link them to phenotypes with greater statistical and biological rigor.

Through contributions from experts across computational biology, biostatistics, machine learning, and systems genetics, this Collection will provide the research community with practical and reproducible workflows for end-to-end multi-omics network analysis. Protocols will emphasize transparent implementation, open-source software, and examples across a range of biological contexts, from complex diseases to developmental biology and toxicology. Ultimately, this Methods Collection aims to accelerate methodological innovation, promote best practices, and empower investigators to extract meaningful biological insights from increasingly complex multi-omics datasets.