Fig. 1: Spatial multi-modal processing workflow and integration using the MAGPIE framework.

a Experimental workflow for spatial multi-omics profiling of rodent lung, consisting of agarose inflation of the lung tissue to facilitate the preservation of tissue integrity, followed by snap-freezing and cryo-sectioning to create consecutive sections, used for spatial transcriptomics and spatial metabolomics, respectively. b MAGPIE computational framework for co-registering same or consecutive section spatial transcriptomics (Visium) and metabolomics (mass spectrometry imaging, MSI) data. The pipeline’s inputs and outputs are in standardised Space Ranger-style and tabular formats to ensure compatibility with other tools. Preprocessing the MSI data produces a data-generated image used for landmark selection and subsequent image co-registration. The output from the pipeline is the MSI data with updated coordinates aligned to the Visium data. c To create a 1:1 mapping between Visium spots and MSI pixels, MAGPIE (by default) expands Visium spot radii, such that there are no gaps between spots, and then aggregates MSI pixels that fall within these spot boundaries. This results in matching observations between modalities and a processed object which can then be read by analysis toolkits, including semla²⁶. d Overview of downstream analysis options once the modalities are aligned into a matching coordinate system. Examples include joint clustering, dimensionality reduction, linking transcriptomic and metabolic changes to histologic information, and gene-peak correlation or multi-omics covariation network analysis. Spatial b.c. spatial barcode, CCF common coordinate framework. Source data are provided as a Source Data file.