IMC datasets for "Spatiotemporal organisation of residual disease in mouse and human BRCA1-deficient mammary tumours and breast cancer"

IMC datasets for "Spatiotemporal organisation of residual disease in mouse and human BRCA1-deficient mammary tumours and breast cancer" This repository contains paired IMC-Visium SpatialData objects stored as Zarr datasets created and analysed in our article: Demeter Túrós, Morgane Decollogny, Astrid Chanfon, Myriam Siffert, Lou Romanens, Jean-Christophe Tille, Olivier Tredan, Intidhar Labidi-Galy, Alberto Valdeolivas, and Sven Rottenberg."Spatiotemporal organisation of residual disease in mouse and human BRCA1-deficient mammary tumours and breast cancer." GitHub: https://github.com/rottenberglab/residual-disease Archive Contents Co-registered IMC and 10x Visium samples of cisplatin-treated mouse mammary tumors. For each sample pair, two Zarr groups are generated: one for IMC and one for Visium. The IMC Zarr contains phenotyped and segmented cells within the main Table of the SpatialData object, while the Visium Zarr stores transcriptional information at the capture spot level. Zarr datasets were prepared using SpatialData (https://spatialdata.scverse.org).To load it, use the following: import spatialdata as sdsdata = sd.read_zarr('your_path') imc.zarrs contain the following data: SpatialData object with:├── Images│ ├── 'SAMPLE_ID_full_image': MultiscaleSpatialImage[cyx]: 8k H&E image│ ├── 'SAMPLE_ID_hires_image': SpatialImage[cyx]: high-resolution image from Space Ranger│ ├── 'SAMPLE_ID_lowres_image': SpatialImage[cyx]: low-resolution image from Space Ranger│ ├── 'panorama_X': SpatialImage[cyx] (1, y, x): IMC grayscale panorama scan│ └── 'panorama_X-ROI_X': SpatialImage[cyx] (54, y, x): IMC regions-of-interest├── Labels│ ├── 'panorama_X-ROI_X-cell_mask': SpatialImage[yx] (y, x): voronoi tessellation cell masks (clipped at a radius of 7 μm) based on nucleus centroids predicted with cellpose│ ├── 'panorama_X-ROI_X-nucleus_mask': SpatialImage[yx] (y, x): Nucleus masks predicted with cellpose│ └── 'panorama_X-ROI_X-mesmer_cell_mask': SpatialImage[yx] (y, x): Cell masks predicted with Mesmer├── Points│ └── 'panorama_X-ROI_X-nucleus_points': DataFrame (2D points): Nucleus centroids├── Shapes│ └── 'panorama_X-ROI_X-cells': GeoDataFrame (2D shapes): Mesmer cell polygons└── Table └── AnnData object with n_obs × n_vars = N × 54 obs: 'label', 'ROI', 'cell_type', 'cp_pt', 'log1p_cp_pt', 'cp_pt_norm' var: 'markers', 'metals', 'masses', 'index' uns: 'roi_normalization_factors', 'spatialdata_attrs' obsm: 'regionprops', 'spatial', 'sum_intensity' obsp: 'spatial_connectivities', 'spatial_distances': AnnData (N, 54) with coordinate systems mapping between the Visium image and the IMC data. visium.zarrs contain the following data: SpatialData object with:├── Images│ ├── 'SAMPLE_ID_full_image': MultiscaleSpatialImage[cyx]: 8k H&E image│ ├── 'SAMPLE_ID_hires_image': SpatialImage[cyx]: high-resolution image from Space Ranger│ ├── 'SAMPLE_ID_lowres_image': SpatialImage[cyx]: low-resolution image from Space Ranger│ ├── 'panorama_X': SpatialImage[cyx] (1, y, x): IMC grayscale panorama scan│ └── 'panorama_X-ROI_X': SpatialImage[cyx] (54, y, x)├── Shapes│ └── 'SAMPLE_ID': GeoDataFrame shape: (N, 2) (2D shapes): └── Table └── AnnData object with n_obs × n_vars = N × M obs: 'in_tissue', 'array_row', 'array_col', 'spot_id', 'region' var: 'gene_ids', 'feature_types', 'genome' uns: 'spatial', 'spatialdata_attrs' obsm: 'spatial': AnnData (N, M) with coordinate systems mapping between the Visium image and the IMC data.