Unsupervised methods in multi-omic data analysis involve techniques for exploring the structure and patterns of the data without prior knowledge of the experimental design. These include cluster analysis and visualisation methods based on ordination. These procedures can reveal meaningful groupings among observations or allow for reducing the complexity of the data by reducing the number of dimensions. Researchers can then use the results from these exploratory techniques in further multi-omic data integration. It’s crucial to properly pre-process the data and choose the appropriate association coefficient when computing these methods, as this has a significant impact on the final outcome.