Spatial Multi-Omics

For a tissue section measured with spatial metabolomics (MALDI / DESI imaging, m/z or compound names) and a second spatial omic on the same section (Visium transcriptomics, spatial proteomics). The metabolite pixels are mapped onto the spots of the second layer, both layers are normalized, and you choose among joint clustering of the spots, cross-omic co-localisation, pseudobulk comparison of both layers, joint pathway analysis and shared spatial programs. Both layers are tabular (peak-picked, filtered to the informative features) and limited to 50 MB each; one section per analysis. For whole-transcriptome matrices, raw imzML or many sections use the MetaboAnalystR package, or reduce the tables first with the Spatial Data Bridge module (Visium converter keeps the most variable genes).

Please upload the two layers (.csv or .txt)
  • Metabolite table: one row per pixel / spot, spot id in the first column, one column per feature (m/z values or compound names).
  • Metabolite metadata (optional): Spot_ID, x, y of the metabolite pixels when they are not the spots of the second layer (e.g. MALDI pixels under Visium spots); leave empty when both tables share the spot ids.
  • Gene / protein table: one row per spot, spot id in the first column, one column per gene or protein (the Visium converter of the Spatial Data Bridge module produces this table from Space Ranger output).
  • Spot metadata: Spot_ID, x, y and a region-of-interest column of the gene / protein spots. Both layers must use the same coordinate system (registered section).
Raw or TIC-scaled metabolite intensities and raw gene counts are expected. Limits: 50 MB per file, at most 100,000 spots, 5,000 metabolite features and 5,000 genes (the most variable genes are kept).
Choose
Choose
Choose
Choose
(auto = first of ROI / group / region / cluster)
MALDI-MSI table
MSI pixel metadata
Visium gene table
Visium spot metadata
One coronal mouse brain section (striatum, section V11L12-038_A1) profiled by Spatial Multimodal Analysis (Vicari et al., Nat Biotechnol 2024): MALDI-FT-ICR imaging of lipids (DHB, positive mode, 3,105 tissue pixels x 600 most variable m/z) and Visium RNA on the same section (2,856 spots x 1,012 most variable genes, counts). The MSI pixel grid was registered onto the Visium coordinates (micrometres) from the tissue outline and the region labels; the regions (cortex, caudoputamen, corpus callosum, ...) come from the authors' Loupe annotation. Source: Mendeley Data 10.17632/w7nw4km7xd.1.