The vast amounts of multi-omics data have brought unprecedented opportunities for biomedical data discovery. However, the complexity and heterogeneity of the data also poses great challenges for fast yet accurate analysis. Multi-omics datasets are usually organized in two board ways, vertical or horizontal, depending on the specific question of interest.

In the vertical setting, multiple technologies are used to generate data from different aspects of the research question, including, but not limited to, genome, epigenome, transcriptome, proteome, metabolome, or microbiome. The major challenges in analyzing these complex multi-layered data modalities include the identification of interactions within and across data modalities, as well as the construction and interpretation of networks. In the horizontal setting, multiple datasets are generated from one or two technologies for a specific research question. These datasets are typically from diverse populations across the world, representing a high degree of real-world biological and technical heterogeneity. The major challenges in analyzing these heterogenous multi-cohort datasets include data integration, meta-analysis, and identification of the most robust signals encompassing data heterogeneity. The vertical and horizontal structures correspond to the complexity and heterogeneity of the multi-omics data.

The aim of this Special Issue is to provide investigators with a platform to share their research relating to leveraging complexity and heterogeneity in multi-omics data for biomedical data discovery, which can be applied for better diagnosis, treatment, prognosis, and prevention of human diseases, in the future era of precision medicine.

Potential topics include but are not limited to the following:

• Algorithms, methods, frameworks, and best practices for multi-omics data analysis, addressing the vertical complexity and horizontal heterogeneity aspects in data analysis
• Methods for identifying interactions between different data modalities
• Benchmarks of network construction methods
• Review of meta-analysis frameworks
• Methods for integrating biomedical imaging data, such as computed tomography or hematoxylin-eosin staining
• Methods for integrating non-imaging data, such as next generation sequencing data
• Methods for merging data from different batches, especially for data with strong batch effects, such as single cell RNA sequencing data
• Application of algorithms, methods, or frameworks for disease diagnosis, treatment, prognosis, and prevention
• Multi-omics profiling and network identification of specific diseases
• Meta-analyses of multi-cohort datasets for specific diseases
• Experimental validation of biomarkers identified from multi-omics data analysis
• Disease diagnosis and prognosis prediction from imaging and non-imaging data analysis
• Clinical applications or validations of findings from multi-omics data analysis