This repository hosts the results of processing example imaging mass cytometry (IMC) data hosted at 10.5281/zenodo.5949116 using the IMC Segmentation Pipeline available at https://github.com/BodenmillerGroup/ImcSegmentationPipeline (DOI: 10.5281/zenodo.6402666). Please refer to https://github.com/BodenmillerGroup/steinbock as alternative processing framework and 10.5281/zenodo.6043600 for the data generated by steinbock. The following files are part of the analysis.zip folder when running the IMC Segmentation Pipeline: cpinp: contains input files for the segmentation pipeline cpout: contains all final output files of the pipeline: cell.csv containing the single-cell features; Experiment.csv containing CellProfiler metadata; Image.csv containing acquisition metadata; Object relationships.csv containing an edge list indicating interacting cells; panel.csv containing channel information; var_cell.csv containing cell feature information; var_Image.csv containing acquisition feature information; images containing the hot pixel filtered multi-channel images and the channel order; masks containing the segmentation masks; probabilities containing the pixel probabilities. histocat: contains single channel .tiff files per acquisition for upload to histoCAT (https://bodenmillergroup.github.io/histoCAT/) crops: contains upscaled image crops in .h5 format for ilastik (https://www.ilastik.org/) training ometiff: contains .ome.tiff files per acquisition, .png files per panorama and additional metadata files per slide ilastik: multi channel images for ilastik pixel classification (_ilastik.full) and their channel order (_ilastik.csv); upscaled multi channel images for ilastik pixel prediction (_ilastik_s2.h5); upscaled 3 channel images containing ilastik pixel probabilities (_ilastik_s2_Probabilities.tiff). The remaining files are part of the root directory: docs.zip: Documentation of the pipeline in markdown format IMCWorkflow.ilp: Ilastik pixel classifier pre-trained on the example data resources.zip: The CellProfiler pipelines and CellProfiler plugins used for the analysis sample_metadata.xlsx: Metadata per sample including the cancer type scripts.zip: Python notebooks used for pre-processing and downloading the example data src.zip: Scripts for the imcsegpipe python package

Imaging mass cytometry (IMC) data was collected on multiple regions of interest (ROIs) from a racially balanced and clinically matched cohort of 57 surgically resected tissues, primarily TNBC, as identified by H&E images. This cohort consisted of 26 self-reported Black American (BA) women and 31 self-reported White American (WA) women. ROIs were selected from both the tumor center and tumor periphery, and were categorized as either immune-rich or immune-poor.

The folder consists of a text file (readme.txt) and three Matlab scripts demonstrating the three simulations in Localization toy problem. (ZIP)

Processed and annotated single-cell data (.RDS) associated with Cho Y et al. manuscript entitled, "Modeling cellular influence delineates functionally relevant cellular neighborhoods in primary and metastatic pancreatic ductal adenocarcinoma".

Raw (.MCD) files are located: https://doi.org/10.5281/zenodo.15601931

Code for analysis is located: https://github.com/DeshpandeLab/Spatial_Influence

ometiff: Imaging Data

Cell Masks: Masks generated with cellprofiler from ilastik segmentation training

cp-output_config: All relevant cellprofiler output and additional configuration files (for example clinical data) necessary to generate the single cell experiments.

IMC Data Objects: Single cell experiment RDS files.

Imaging Mass Cytometry (IMC) is a technology that enables comprehensive analysis of cellular phenotypes at the tissue level. We performed a multi-parameter characterization of structural and immune cell populations in psoriatic skin and synovial tissue samples aimed at characterizing immune cell differences in patients with psoriasis, psoriatic arthritis (PsA). A panel of 33 antibodies was used to stain selected immune and structural cell populations. IMC data were segmented into single cells based on combinations of antibody stains. Single cells were then clustered into cell categories based on pre-specified markers. The spatial relationships of different cell populations were assessed using neighborhood analysis. Among all cell types in the skin and synovium, lymphoid cells accounted for the most prevalent cell type. T cells and macrophages were the most prevalent immune cell type in the synovium and B cells and NK cells were also identified. Neighborhood analysis showed high correlation between synovial T cells, B cells, macrophages, dendritic cells and neutrophils suggesting spatial organization. Innate and adaptive immune cells can be reliably identified using IMC in skin and synovium. Inter-patient heterogeneity exists in tissue cell populations. IMC provides opportunities for exploring in depth underlying immunological mechanisms driving psoriasis and PsA.

Localization accuracy from measurements of Simulation 1.

Single cell protein expression extracted from segmented IMC image

Recent studies have provided insights into the pathology and immune response to coronavirus disease 2019 (COVID-19). However thorough interrogation of the interplay between infected cells and the immune system at sites of infection is lacking. We use high parameter imaging mass cytometry targeting the expression of 36 proteins, to investigate at single cell resolution, the cellular composition and spatial architecture of human acute lung injury including SARS-CoV-2. This spatially resolved, single-cell data unravels the disordered structure of the infected and injured lung alongside the distribution of extensive immune infiltration. Neutrophil and macrophage infiltration are hallmarks of bacterial pneumonia and COVID-19, respectively. We provide evidence that SARS-CoV-2 infects predominantly alveolar epithelial cells and induces a localized hyper-inflammatory cell state associated with lung damage. By leveraging the temporal range of COVID-19 severe fatal disease in relation to the time of symptom onset, we observe increased macrophage extravasation, mesenchymal cells, and fibroblasts abundance concomitant with increased proximity between these cell types as the disease progresses, possibly as an attempt to repair the damaged lung tissue. This spatially resolved single-cell data allowed us to develop a biologically interpretable landscape of lung pathology from a structural, immunological and clinical standpoint. This spatial single-cell landscape enabled the pathophysiological characterization of the human lung from its macroscopic presentation to the single-cell, providing an important basis for the understanding of COVID-19, and lung pathology in general.

Raw (.MCD) files associated with Cho Y et al. manuscript entitled, "Modeling cellular influence delineates functionally relevant cellular neighborhoods in primary and metastatic pancreatic ductal adenocarcinoma".

Processed and annotated single-cell data (.RDS) are located: https://doi.org/10.5281/zenodo.15596960

Code for analysis is located: https://github.com/DeshpandeLab/Spatial_Influence

This repository hosts the results of processing example imaging mass cytometry (IMC) data hosted at zenodo.org/record/5949116 using the steinbock framework available at github.com/BodenmillerGroup/steinbock. Please refer to steinbock.sh for how these data were generated from the raw data.

The following files are part of this repository:

• panel.csv: contains channel information regarding the used antibodies in steinbock format
• img.zip: contains hot pixel filtered multi-channel images derived from the IMC raw data. One file per acquisition is generated
• images.csv: contains metadata per acquisition
• pixel_classifier.ilp: ilastik pixel classifier (same as the one in zenodo.org/record/6043544)
• ilastik_crops.zip: image crops on which the ilastik classifier was trained (same as the ones in zenodo.org/record/6043544)
• ilastik_img.zip: contains multi-channel images (one per acquisition) in .h5 format for ilastik pixel classification
• ilastik_probabilities.zip: 3 channel images containing the pixel probabilities after pixel classification
• masks_ilastik.zip: segmentation masks derived from the ilastik pixel probabilities using the cell_segmentation.cppipe pipeline
• masks_deepcell.zip: segmentation masks derived by deepcell segmentation
• intensities.zip: Contains one .csv file per acquisition. Each file contains single-cell measures of the mean pixel intensity per cell and channel based on the files in img.zip and masks_deepcell.zip.
• regionprops.zip: Contains one .csv file per acquisition. Each file contains single-cell measures of the morphological features and location of cells based on masks_deepcell.zip.
• neighbors.zip: Contains one .csv file per acquisition. Each file contains an edge list of cell IDs indicating cells in close proximity based on masks_deepcell.zip.
• ome.zip: contains .ome.tiff files derived from img.zip; one file per acquisition
• histocat.zip: contains single-channel .tiff files with segmentation masks derived from masks_deepcell.zip for upload to histoCAT (bodenmillergroup.github.io/histoCAT)
• cells.csv: contains intensity and regionprop measurements of all cells
• cells_csv.zip: contains intensity and regionprop measurements of all cells per acquisition
• cells.fcs: contains intensity and regionprop measurements of all cells in fcs format
• cells_fcs.zip: contains intensity and regionprop measurements of all cells per acquisition in fcs format
• cells.h5ad: contains intensity, regionprop and neighbor measurements of all cells in anndata format
• cells_h5ad: contains intensity regionprop and neighbor measurements of all cells per acquisition in anndata format
• graphs.zip: contains spatial object graphs in .graphml format; one file per acquisition