IRIS: time-structured manifold projections
This work provides a new visualization tool for researchers working with time-series high-dimensional biomedical data, enabling clearer understanding of dynamic processes.
This paper introduces IRIS, a novel Manifold Learning algorithm designed to visualize high-dimensional biomedical data that is generated temporally. IRIS structures layouts chronologically while preserving manifold topology, addressing the limitation of existing methods like t-SNE and UMAP which cannot incorporate time-ordering.
High-dimensional biomedical data, such as cell-by-gene matrices, are increasingly generated temporally. However, Manifold Learning algorithms, like t-SNE and UMAP, cannot incorporate time-ordering in their layouts, obfuscating the dynamics of cell types or other classes. As a solution, we present IRIS, a new Manifold Learning algorithm that structures layouts both chronologically and by manifold topology. IRIS can visualize a wide range of dynamic biomedical data, including scRNA-seq, comparative metagenomics, and literature.