Modern omics technologies generate large-scale gene expression datasets across thousands of genes and samples. While these data contain rich biological signals, it remains challenging to systematically identify meaningful patterns without prior assumptions. To address this, we developed singleCellHaystack, a bioinformatics method for detecting structured gene expression patterns in high-dimensional data, including single-cell and spatial omics (Vandenbon and Diez, Nature Communications, 2020). The identified patterns can be mapped back onto tissues or samples, enabling the discovery of biologically relevant gene activities across conditions. The figure shows representative examples from mouse liver (left), human intestine (top right), and mouse brain (bottom right).

Spatial transcriptomics technologies enable high-resolution mapping of gene expression within tissues, but experiments remain costly and data processing requires substantial computational expertise. As a result, effective reuse and exploration of existing datasets is often limited. To address this, we developed DeepSpaceDB, a database and analysis platform that collects, standardizes, and integrates publicly available spatial transcriptomics data (Honcharuk et al., Nucleic Acids Res. 2026). DeepSpaceDB enables users to explore gene expression patterns, compare samples, and perform interactive analyses through an intuitive interface, without requiring advanced bioinformatics skills. The figure illustrates the data processing pipeline (left), the integration of more than 3,000 spatial datasets (center), and representative interactive analysis functions (right).