Preprocessed h5ad files of Xenium spatial transcriptomics performed on atherosclerotic human carotid artery samples

Xenium_final_dataset.zip Xenium runs were performed with two panels of genes. The panels consist of a core list present in both panels (comprising 111 central genes for cell type annotations and key processes) as well as individual genes for the panels, resulting in a cell-defining (panel 1) and disease associated (panel 2) panel. In total 548 unique genes have been sequenced with both panels. Sequencing with panel 1 and panel 2 genes was performed on consecutive FFPE sections for each sample, thereby introducing a 5 µm shift in the morphology of the mounted Xenium slides across the panels. Xenium_final_dataset.zip contains the Anndata objects for count matrices obtained with panel 1 and panel 2 genes. In addition, the Anndata objects contain patient metadata variables. Description of .obs columns is provided in the Xenium_column_descriptions.md file.IMPORTANT: THE COLUMNS [x_um,y_um] IN ANNDATA.OBS, WHICH CONTAIN THE CELL CENTER COORDINATES IN MICROMETERS ARE INCORRECT IN THIS VERSION DUE TO A BUG IN THE CODE. TO CALCULATE THE CORRECT VALUES IN MICROMETERS, MULTIPLY THE COLUMNS [x_pixel, y_pixel] by 0.2125!Subregion_coordinates.zip The annotation of sample subregions is a histological analysis, using established Oxford Plaque Studies and AHA criteria and was performed by an experienced cardiovascular pathologist. The borders of the annotated subregions were initially determined on the HE stained slides by a pathology expert, then transferred to the DAPI-stained slides manually by drawing the determined subregion borders by hand in Xenium Explorer. The manually anootated subregion coordinates for individual samples are contained in the Subregion_coordinates.zip file. Note, that the "_D" suffix corresponds to plaque, whereas the "_H" suffix to control samples. subregion_area_dict.pickle Border coordinates were extracted from Xenium Explorer and converted to polygons, which then were used to assign individual cells to their respective subregions and to calculate the subregion areas. Subregion polygons and areas of individual samples, along with whole sample areas are contained in subregion_area_dict.pickle, in a Python pickled dictionary format. Detailed description of preprocessing steps can be found in our manuscript.