| Title: | Collection of publicly available imaging mass cytometry (IMC) datasets |
|---|---|
| Description: | The imcdatasets package provides access to publicly available IMC datasets. IMC is a technology that enables measurement of > 40 proteins from tissue sections. The generated images can be segmented to extract single cell data. Datasets typically consist of three elements: a SingleCellExperiment object containing single cell data, a CytoImageList object containing multichannel images and a CytoImageList object containing the cell masks that were used to extract the single cell data from the images. |
| Authors: | Nicolas Damond [aut] (ORCID: <https://orcid.org/0000-0003-3027-8989>), Nathan Steenbuck [aut, cre] (ORCID: <https://orcid.org/0000-0003-2712-5359>), Nils Eling [ctb] (ORCID: <https://orcid.org/0000-0002-4711-1176>), Fischer Jana [ctb] (ORCID: <https://orcid.org/0000-0001-9047-8400>) |
| Maintainer: | Nathan Steenbuck <[email protected]> |
| License: | GPL-3 + file LICENSE |
| Version: | 1.21.3 |
| Built: | 2026-10-10 11:19:16 UTC |
| Source: | https://github.com/bioc/imcdatasets |
Obtain the Damond_2019_Pancreas dataset, which consists of three data objects: single cell data, multichannel images and cell segmentation masks. The data was obtained by imaging mass cytometry (IMC) of human pancreas sections from donors with type 1 diabetes.
Damond_2019_Pancreas( data_type = c("sce", "spe", "images", "masks"), full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )Damond_2019_Pancreas( data_type = c("sce", "spe", "images", "masks"), full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
full_dataset |
if FALSE (default), a subset corresponding to 100 images
is returned. If TRUE, the full dataset (corresponding to 845 images) is
returned. Due to memory space limitations, this option is only available for
single cell data and masks, not for |
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset from Damond et al. (2019):
images contains a hundred 38-channel
images in the form of a CytoImageList class object.
masks contains the cell segmentation
masks associated with the images, in the form of a
CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a
SingleCellExperiment. This represents a total of 252,059
cells x 38 channels.
spe same single cell data as for sce, but in the
SpatialExperiment format.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment and
SpatialExperiment objects. Mapping at the image level can be
performed with the image_name or image_number variables.
Mapping between cell segmentation masks and single cell data is performed
with the cell_number variable, the values of which correspond to the
intensity values of the masks object. For practical
examples, please refer to the "Accessing IMC datasets" vignette.
This dataset is a subset of the complete Damond et al. (2019) dataset
comprising the data from three pancreas donors at different stages of type 1
diabetes (T1D). The three donors present clearly diverging characteristics
in terms of cell type composition and cell-cell interactions, which makes
this dataset ideal for benchmarking spatial and neighborhood analysis
algorithms. If full_dataset = TRUE, the full dataset (845 images from
12 patients) is returned. This option is not available for multichannel
images.
The assay slots of the SingleCellExperiment and
SpatialExperiment objects contain three assays:
counts contains raw mean ion counts per cell.
exprs contains arsinh-transformed counts, with cofactor 1.
quant_norm contains counts censored at the 99th percentile
and scaled 0-1.
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment / SpatialExperiment
objects.
The cell-associated metadata are stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. These metadata include cell types (in
colData(sce)$cell_type) and broader cell categories, such as
"immune" or "islet" cells (in colData(sce)$cell_category). In
addition, for cells located inside pancreatic islets, the islet they belong
to is indicated in colData(sce)$islet_parent. For cells not located
in islets, the "islet_parent" value is set to 0 but the spatially closest
islet can be identified with colData(sce)$islet_closest.
The donor-associated metadata are also stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. For instance, the donors' IDs can be retrieved with
colData(sce)$patient_id and the donors' disease stage can be obtained
with colData(sce)$patient_stage.
Neighborhood information, defined here as cells that are localized next to
each other, is stored as a SelfHits object in the colPairs
slot of the SingleCellExperiment and SpatialExperiment
objects.
The three donors in the subset present the following characteristics:
6126 is a non-diabetic donor, with large islets containing
many beta cells, severe infiltration of the exocrine pancreas with
myeloid cells but limited infiltration of islets.
6414 is a donor with recent T1D onset (shortly after
diagnosis) showing partial beta cell destruction and mild infiltration
of islets with T cells.
6180 is a donor with long-duration T1D (11 years after
diagnosis), showing near-total beta cell destruction and limited immune
cell infiltration in both the islets and the pancreas.
For information about other donors in the full dataset, please refer to the Damond et al. publication.
Dataset versions: a version argument can be passed to the function to
specify which dataset version should be retrieved.
`v0`: original version (Bioconductor <= 3.15).
`v1`: consistent object formatting across datasets.
File sizes:
`images`: size in memory = 7.4 Gb, size on disk = 1.7 Gb.
`masks`: size in memory = 200 Mb, size on disk = 8.2 Mb.
`sce`: size in memory = 353 Mb, size on disk = 204 Mb.
`spe`: size in memory = 372 Mb, size on disk = 205 Mb.
`sce_full`: size in memory = 2.4 Gb, size on disk = 1.5 Gb.
`spe_full`: size in memory = 2.5 Gb, size on disk = 1.5 Gb.
`masks_full`: size in memory = 1.4 Gb,
size on disk = 60 Mb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
Original source: Damond et al. (2019): https://doi.org/10.1016/j.cmet.2018.11.014
Original link to raw data, also containing the entire dataset: https://data.mendeley.com/datasets/cydmwsfztj/2
A SingleCellExperiment object with single cell data, a SpatialExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Nicolas Damond
Damond N et al. (2019). A Map of Human Type 1 Diabetes Progression by Imaging Mass Cytometry. Cell Metab 29(3), 755-768.
# Load single cell data sce <- Damond_2019_Pancreas(data_type = "sce") print(sce) # Display metadata Damond_2019_Pancreas(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- Damond_2019_Pancreas(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- Damond_2019_Pancreas(data_type = "sce") print(sce) # Display metadata Damond_2019_Pancreas(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- Damond_2019_Pancreas(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))
Obtain the HochSchulz_2022_Melanoma dataset, which is composed of two panels (rna and protein) that were acquired on consecutive sections. Each dataset (panel) is composed of three data objects: single cell data, multichannel images and cell segmentation masks. The data was obtained by imaging mass cytometry (IMC) of a tissue microarray (TMA) with multiple cores of formalin-fixed paraffin-embedded (FFPE) tissue from 69 patients with metastatic melanoma.
HochSchulz_2022_Melanoma( data_type = c("sce", "spe", "images", "masks"), panel = "rna", full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )HochSchulz_2022_Melanoma( data_type = c("sce", "spe", "images", "masks"), panel = "rna", full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
panel |
which panel should be returned? Can be set to "rna" (default) or "protein". |
full_dataset |
if FALSE (default), a subset corresponding to the 50
images containing the most B cells is returned. If TRUE, the full dataset
(corresponding to 166 images) is returned. Due to memory space limitations,
this option is only available for single cell data and masks, not for
|
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset from Hoch, Schulz et al. (2022):
images contains fifty 38-channel
images in the form of a CytoImageList class object.
masks contains the cell segmentation
masks associated with the images, in the form of a
CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a
SingleCellExperiment object.
spe same single cell data as for sce, but in the
SpatialExperiment format.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment and
SpatialExperiment objects. Mapping at the image level can be
performed with the image_name or image_number variables.
Mapping between cell segmentation masks and single cell data is performed
with the cell_number variable, the values of which correspond to the
intensity values of the masks object. For practical
examples, please refer to the "Accessing IMC datasets" vignette.
The assay slots of the SingleCellExperiment and
SpatialExperiment objects contain three assays:
counts contains raw mean ion counts per cell.
exprs contains arsinh-transformed counts, with cofactor 1.
scaled_counts contains scaled counts.
scaled_exprs contains scaled asinh-transformed counts.
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment / SpatialExperiment
objects.
The cell-associated metadata are stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. These metadata include various information about cells, milieu,
samples, and patients. For instance, cell types can be retrieved with
colData(sce)$cell_type and cell clusters with
colData(sce)$cell_cluster.
Neighborhood information, defined here as cells that are localized next to
each other, is stored as a SelfHits object in the colPairs
slot of the SingleCellExperiment and SpatialExperiment
objects.
For more information, please refer to the Hoch, Schulz, et al. publication.
Dataset versions: a version argument can be passed to the function to
specify which dataset version should be retrieved.
`v1`: first published version
File sizes:
`images_rna`: size in memory = 13.9 Gb,
size on disk = 954 Mb.
`masks_rna`: size in memory = 347 Mb,
size on disk = 11 Mb.
`sce_rna`: size in memory = 774 Mb,
size on disk = 401 Mb.
`masks_full_rna`: size in memory = 1.1 Gb,
size on disk = 30 Mb.
`sce_full_rna`: size in memory = 2.0 Gb,
size on disk = 1.1 Gb.
`images_protein`: size in memory = 16.8 Gb,
size on disk = 1.2 Gb.
`masks_protein`: size in memory = 374 Mb,
size on disk = 12 Mb.
`sce_protein`: size in memory = 856 Mb,
size on disk = 531 Mb.
`masks_full_protein`: size in memory = 1.2 Gb,
size on disk = 35 Mb.
`sce_full_protein`: size in memory = 2.2 Gb,
size on disk = 1.4 Gb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
Original source: Hoch, Schulz et al. (2022): https://doi.org/10.1126/sciimmunol.abk1692
Original link to raw data: https://doi.org/10.5281/zenodo.5994136.
A SingleCellExperiment object with single cell data, a SpatialExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Nicolas Damond
Hoch, Schulz et al. (2022). Multiplexed imaging mass cytometry of the chemokine milieus in melanoma characterizes features of the response to immunotherapy Sci Immunol 7(70):eabk1692.
# Load single cell data sce <- HochSchulz_2022_Melanoma(data_type = "sce") print(sce) # Display metadata HochSchulz_2022_Melanoma(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- HochSchulz_2022_Melanoma(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- HochSchulz_2022_Melanoma(data_type = "sce") print(sce) # Display metadata HochSchulz_2022_Melanoma(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- HochSchulz_2022_Melanoma(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))
Obtain the IMMUcan_2022_CancerExample dataset, which consists of three data objects: single cell data, multichannel images and cell segmentation masks. Data were obtained by imaging mass cytometry (IMC) of sections of 4 patients with different tumor indications.
IMMUcan_2022_CancerExample( data_type = c("sce", "spe", "images", "masks"), version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )IMMUcan_2022_CancerExample( data_type = c("sce", "spe", "images", "masks"), version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset used in the IMC data analysis book
images contains 14 multichannel images, each
containing 50 channels, in the form of a CytoImageList class
object.
masks contains the cell segmentation masks associated
with the images, in the form of a CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a SingleCellExperiment
object. Single cell data can also be retrieved as a
SpatialExperiment object. This represents a total of 46,825
cells x 40 channels.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment
object. Mapping at the image level can be performed with the
sample_id or image_name variables. Mapping between cell
segmentation masks and single cell data is performed with the
cell_number variable, the values of which correspond to the
intensity values of the masks object. For practical examples, please
refer to the "Accessing IMC datasets" vignette.
This imaging mass cytometry dataset serves as an example to demonstrate downstream analysis tools including spatial data analysis. The data was generated as part of the Integrated iMMUnoprofiling of large adaptive CANcer patient cohorts (IMMUcan) project (immucan.eu) using the Hyperion imaging system.
Relevant entries to the colData slot are as follows:
sample_id image name.
cell_number cell identifier.
width_px width of the image.
height_px height of the image.
patient_id patient identifier.
ROI region of interest identifier.
indication cancer type.
cell_labels labels of manually labelled cells.
cell_type cell type as defined by classification.
spatial_community identifiers of each spatial tumor or
non-tumor community
cn_celltypes cellular neighborhoods as defined by
clustering cells based on the frequency of neighboring cell types.
cn_expression cellular neighborhoods as defined by
clustering cells based on the mean expression of neighboring cells
lisa_clusters cellular neighborhoods as detected by the
lisaClust package.
spatial_context spatial contexts defined in
cn_celltype.
spatial_context_filtered filtered spatial context
identifiers.
patch_id identifier of the spatial tumor patch.
cell_x spatial x coordinate.
cell_y spatial y coordinate.
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment object.
The assay slot of the SingleCellExperiment object
contains two assays:
counts: mean ion counts per cell
exprs: arsinh-transformed counts per cell, with cofactor 1.
The colPair slot of the SingleCellExperiment object
contains the following spatial object graphs:
neighborhood steinbock generated graph.
knn_interaction_graph 20-nearest neighbor graph.
expansion_interaction_graph expansion graph using a threshold
of 20.
delaunay_interaction_graph interaction graph constructed by
delaunay triangulation.
knn_spatialcontext_graph 40-nearest neighbor graph.
File sizes:
`images`: size in memory = 1.5 Gb, size on disk = 786 Mb.
`masks`: size in memory = 19 Mb, size on disk = 1.2 Mb.
`sce`: size in memory = 182 Mb, size on disk = 82 Mb.
`spe`: size in memory = 183 Mb, size on disk = 81 Mb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
A SingleCellExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Nils Eling
# Load single cell data sce <- IMMUcan_2022_CancerExample(data_type = "sce") print(sce) # Display metadata IMMUcan_2022_CancerExample(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- IMMUcan_2022_CancerExample(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- IMMUcan_2022_CancerExample(data_type = "sce") print(sce) # Display metadata IMMUcan_2022_CancerExample(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- IMMUcan_2022_CancerExample(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))
Obtain the JacksonFischer_2020_BreastCancer dataset, which consists of three data objects: single cell data, multichannel images and cell segmentation masks. The data was obtained by imaging mass cytometry (IMC) of tumour tissue from patients with breast cancer.
JacksonFischer_2020_BreastCancer( data_type = c("sce", "spe", "images", "masks"), full_dataset = FALSE, cohort = "Basel", version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )JacksonFischer_2020_BreastCancer( data_type = c("sce", "spe", "images", "masks"), full_dataset = FALSE, cohort = "Basel", version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
full_dataset |
if FALSE (default), a subset corresponding to 100 images
is returned. If TRUE, the full dataset is returned, including both "Basel"
and "Zurich" cohorts. Due to memory space limitations, this option is only
available for single cell data and masks, not for
|
cohort |
which patient cohort should be returned? Can be set to "Basel"
(default) or "Zurich". Ignored if |
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset from Jackson, Fischer et al. (2020):
images contains a hundred 42-channel
images in the form of a CytoImageList class object.
masks contains the cell segmentation
masks associated with the images, in the form of a
CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a
SingleCellExperiment. This represents a total of 285,851
cells x 42 channels.
spe same single cell data as for sce, but in the
SpatialExperiment format.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment and
SpatialExperiment objects. Mapping at the image level can be
performed with the image_name variable. Mapping between cell
segmentation masks and single cell data is performed with the
cell_number variable, the values of which correspond to the intensity
values of the masks object. For practical examples, please refer
to the "Accessing IMC datasets" vignette.
This dataset is a subset of the complete Jackson, Fischer et al. (2020)
dataset comprising the data from tumour tissue from 100 patients with breast
cancer (one image per patient). By default, data from the "Basel" cohort are
returned. By setting cohort = "Zurich", data from the "Zurich"
cohort, corresponding to images and associated data from 72 patients, are
returned. For details about the patient cohorts, refer to the publication.
If full_dataset = TRUE, the full dataset is returned (including both
"Basel" and "Zurich" patient cohorts). This option is not available for
multichannel images.
The assay slot of the SingleCellExperiment object
contains three assays:
counts contains mean ion counts per cell.
exprs contains arsinh-transformed counts, with cofactor 1.
quant_norm contains quantile-normalized counts (0 to 1,
99th percentile).
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment and SpatialExperiment
objects.
The cell-associated metadata are stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. These metadata include clusters (in
colData(sce)$cell_cluster_phenograph) and metaclusters (in
colData(sce)$cell_metacluster), as well as spatial information (e.g.,
cell areas are stored in colData(sce)$cell_area).
The clinical data are also stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. For instance, the tumor grades can be retrieved with
colData(sce)$tumor_grade.
Dataset versions: a version argument can be passed to the function to
specify which dataset version should be retrieved.
`v0`: original version (Bioconductor <= 3.15).
`v1`: consistent object formatting across datasets.
`v2`: added full datasets and Zurich cohort.
File sizes:
`images_basel`: size in memory = 19 Gb,
size on disk = 2.0 Gb.
`masks_basel`: size in memory = 433 Mb,
size on disk = 10 Mb.
`sce_basel`: size in memory = 513 Mb,
size on disk = 270 Mb.
`images_zurich`: size in memory = 6.0 Gb,
size on disk = 724 Mb.
`masks_zurich`: size in memory = 137 Mb,
size on disk = 3.4 Mb.
`sce_zurich`: size in memory = 188 Mb,
size on disk = 105 Mb.
`masks_full`: size in memory = 2.1 Gb,
size on disk = 10 Mb.
`sce_full`: size in memory = 2.2 Gb,
size on disk = 1.2 Gb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
Original source: Jackson, Fischer et al. (2020): https://doi.org/10.1038/s41586-019-1876-x
Original link to raw data, containing the entire dataset: https://doi.org/10.5281/zenodo.3518284
A SingleCellExperiment object with single cell data, a SpatialExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Nicolas Damond
Jackson, Fischer et al. (2020). The single-cell pathology landscape of breast cancer. Nature 578(7796), 615-620.
# Load single cell data sce <- JacksonFischer_2020_BreastCancer(data_type = "sce") print(sce) # Display metadata JacksonFischer_2020_BreastCancer(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- JacksonFischer_2020_BreastCancer(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- JacksonFischer_2020_BreastCancer(data_type = "sce") print(sce) # Display metadata JacksonFischer_2020_BreastCancer(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- JacksonFischer_2020_BreastCancer(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))
Summary information for all available datasets in the imcdatasets package.
listDatasets()listDatasets()
Each dataset contains single-cell data, multichannel images and cell segmentation masks.
A DataFrame where each row corresponds to a dataset, containing the fields:
FunctionCall, the R function call required to construct the
dataset.
Species, species of origin.
Tissue, the tissue that was imaged.
NumberOfCells, the total number of cells in the dataset.
NumberOfImages, the total number of images in the dataset.
NumberOfChannels, the number of channels per image.
Reference, a Markdown-formatted citation to
scripts/ref.bib in the imcdatasets installation directory.
listDatasets()listDatasets()
Obtain the Meyer_2025_TripleNegativeBreastCancer dataset. The dataset is composed of three data objects: single cell data, multichannel images and cell segmentation masks. The data was obtained by imaging mass cytometry (IMC) of a tissue microarray (TMA) with multiple cores of formalin-fixed paraffin-embedded (FFPE) tissue from 215 patients with triple-negative breast cancer.
Meyer_2025_TripleNegativeBreastCancer( data_type = c("sce", "spe", "images", "masks"), full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )Meyer_2025_TripleNegativeBreastCancer( data_type = c("sce", "spe", "images", "masks"), full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
full_dataset |
if FALSE (default), a subset corresponding to
to 125 images from 60 patients sampled across proposed patient groups is
returned. This includes all images visually presented in the publication.
If TRUE, the full dataset (corresponding to 450 images) is returned.
Due to memory space limitations, this option is only available for single
cell data and masks, not for |
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset from Meyer et al. (2025):
images contains 125 39-channel
images in the form of a CytoImageList class object.
masks contains the cell segmentation
masks associated with the images, in the form of a
CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a
SingleCellExperiment object.
spe same single cell data as for sce, but in the
SpatialExperiment format.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment and
SpatialExperiment objects. Mapping at the image level can be
performed with the image_name or image_number variables.
Mapping between cell segmentation masks and single cell data is performed
with the cell_number variable, the values of which correspond to the
intensity values of the masks object. For practical
examples, please refer to the "Accessing IMC datasets" vignette.
The assay slots of the SingleCellExperiment and
SpatialExperiment objects contain three assays:
counts contains raw mean ion counts per cell.
exprs contains arsinh-transformed counts, with cofactor 1.
min_max contains 0-1 normalized .
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment / SpatialExperiment
objects.
The cell-associated metadata are stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. These metadata include various information about cells, tumors
and patients. For instance, cell metacluster can be retrieved with
colData(sce)$cell_metacluster and patient groups with
colData(sce)$patient_patientgroup.
Neighborhood information, defined here as cells that are localized next to
each other, is stored as a SelfHits object in the colPairs
slot of the SingleCellExperiment and SpatialExperiment
objects.
For more information, please refer to the Meyer et al. publication.
Dataset versions: a version argument can be passed to the function to
specify which dataset version should be retrieved.
`v1`: first published version
File sizes:
`images`: size in memory = 20.9 Gb,
size on disk = 1.6 Gb.
`masks`: size in memory = 269 Mb,
size on disk = 8 Mb.
`masks_full`: size in memory = 942 Mb,
size on disk = 29 Mb.
`sce`: size in memory = 451 Mb,
size on disk = 241 Mb.
`sce_full`: size in memory = 1.6 Gb,
size on disk = 866 Mb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
Original source: Meyer et al. (2025): https://doi.org/10.1016/j.ccell.2025.06.019
Original link to raw data: https://zenodo.org/records/15304181.
A SingleCellExperiment object with single cell data, a SpatialExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Lasse Meyer
Meyer, Jackson et al. (2025). A stratification system for breast cancer based on basoluminal tumor cells and spatial tumor architecture Cancer Cell 43(9):1637–1655.e9.
# Load single cell data sce <- Meyer_2025_TripleNegativeBreastCancer(data_type = "sce") print(sce) # Display metadata Meyer_2025_TripleNegativeBreastCancer(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- Meyer_2025_TripleNegativeBreastCancer(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- Meyer_2025_TripleNegativeBreastCancer(data_type = "sce") print(sce) # Display metadata Meyer_2025_TripleNegativeBreastCancer(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- Meyer_2025_TripleNegativeBreastCancer(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))
Obtain the SteenbuckDamond_2026_T1D dataset, which is composed of two panels (Islet + Immune) that were acquired on two consecutive 4 µm sections. Each dataset (panel) is composed of three data objects: single cell data, multichannel images and cell segmentation masks. The data was obtained by imaging mass cytometry (IMC) measuring formalin-fixed paraffin-embedded (FFPE) tissue from 88 individuals along the disease progression of T1D. It contains, both Controls, single and multi-auto-antibody positive T1D donors (sAAb+/mAAb+), and donors with recent and long-standing T1D (Onset/LD).
SteenbuckDamond_2026_T1D( data_type = c("sce", "spe", "images", "masks"), panel = "islet", full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )SteenbuckDamond_2026_T1D( data_type = c("sce", "spe", "images", "masks"), panel = "islet", full_dataset = FALSE, version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
panel |
which panel should be returned? Can be set to "islet" (default) or "immune". |
full_dataset |
if FALSE (default), a subset of 120 informative images is returned.
These are sampled along the disease progression of T1D.
It also contains insulitic images, i.e. T-cell infiltration of islets.
If TRUE, the full dataset (corresponding to 7024 images) is returned.
Due to memory space limitations, this option is only available for single
cell data and masks, not for |
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset from Steenbuck, Damond et al. (2026):
images contains 120 47-channel
images in the form of a CytoImageList class object.
masks contains the cell segmentation
masks associated with the images, in the form of a
CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a
SingleCellExperiment object.
spe same single cell data as for sce, but in the
SpatialExperiment format.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment and
SpatialExperiment objects. Mapping at the image level can be
performed with the image_name or image_number variables.
Mapping between cell segmentation masks and single cell data is performed
with the cell_number variable, the values of which correspond to the
intensity values of the masks object. For practical
examples, please refer to the "Accessing IMC datasets" vignette.
The assay slots of the SingleCellExperiment and
SpatialExperiment objects contain three assays:
raw contains raw mean ion counts per cell.
exprs contains arsinh-transformed counts, with cofactor 1.
scaled contains scaled asinh-transformed raw counts.
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment / SpatialExperiment
objects.
The cell-associated metadata are stored in the colData of the
SingleCellExperiment and SpatialExperiment
objects. These metadata include various information about cells, milieu,
samples, and patients. For instance, cell types can be retrieved with
colData(sce)$cell_type and immune cell clusters with
colData(sce)$cell_cluster_scaled. The latter for the Immune panel only.
Neighborhood information, defined here as cells that are localized next to
each other, is stored as a SelfHits object in the colPairs
slot of the SingleCellExperiment and SpatialExperiment
objects.
For more information, please refer to the Steenbuck, Damond, et al. publication.
Dataset versions: a version argument can be passed to the function to
specify which dataset version should be retrieved.
`v1`: first published version
File sizes:#'
`images_immune`: size in memory = 6.3 Gb,
size on disk = 650.3 Mb.
`masks_immune`: size in memory = 138 Mb,
size on disk = 4.7 Mb.
`sce_immune`: size in memory = 442 Mb,
size on disk = 217.1 Mb.
`masks_full_immune`: size in memory = 8.6 Gb,
size on disk = 337.3 Mb.
`sce_full_immune`: size in memory = 21.5 Gb,
size on disk = 13.0 Gb.
`images_islet`: size in memory = 3.8 Gb,
size on disk = 738.8 Mb.
`masks_islet`: size in memory = 83 Mb,
size on disk = 2.8 Mb.
`sce_islet`: size in memory = 277 Mb,
size on disk = 128.5 Mb.
`masks_full_islet`: size in memory = 5.1 Gb,
size on disk = 194.8 Mb.
`sce_full_islet`: size in memory = 12.8 Gb,
size on disk = 6.8 Gb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
Original source: Steenbuck, Damond, et al. (2026): https://doi.org/10.1038/s42255-026-01559-z
Original link to raw data: https://doi.org/10.5281/zenodo.14968076.
A SingleCellExperiment object with single cell data, a SpatialExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Nathan Steenbuck
Steenbuck, N., Damond, N., et al. (2026). Imaging mass cytometry reveals functional and immunological changes during type 1 diabetes progression in human pancreata. Nat Metab https://doi.org/10.1038/s42255-026-01559-z
# Load single cell data sce <- SteenbuckDamond_2026_T1D(data_type = "sce") print(sce) # Display metadata SteenbuckDamond_2026_T1D(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- SteenbuckDamond_2026_T1D(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- SteenbuckDamond_2026_T1D(data_type = "sce") print(sce) # Display metadata SteenbuckDamond_2026_T1D(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- SteenbuckDamond_2026_T1D(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))
Obtain the Zanotelli_2020_Spheroids dataset, which consists of three data objects: single cell data, multichannel images and cell segmentation masks. The data were obtained by imaging mass cytometry (IMC) of sections of 3D spheroids generated from different cell lines.
Zanotelli_2020_Spheroids( data_type = c("sce", "spe", "images", "masks"), version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )Zanotelli_2020_Spheroids( data_type = c("sce", "spe", "images", "masks"), version = "latest", metadata = FALSE, on_disk = FALSE, h5FilesPath = NULL, force = FALSE )
data_type |
type of object to load, 'images' for multichannel images or
'masks' for cell segmentation masks. Single cell data are retrieved using
either 'sce' for the |
version |
dataset version. By default, the latest version is returned. |
metadata |
if FALSE (default), the data object selected in
|
on_disk |
logical indicating if images in form of
HDF5Array objects (as .h5 files) should be stored on disk
rather than in memory. This setting is valid when downloading |
h5FilesPath |
path to where the .h5 files for on disk representation
are stored. This path needs to be defined when |
force |
logical indicating if images should be overwritten when files with the same name already exist on disk. |
This is an Imaging Mass Cytometry (IMC) dataset from Zanotelli et al. (2020), consisting of three data objects:
images contains 517 multichannel images, each containing 51
channels, in the form of a CytoImageList class object.
masks contains the cell segmentation
masks associated with the images, in the form of a
CytoImageList class object.
sce contains the single cell data extracted from the
multichannel images using the cell segmentation masks, as well as the
associated metadata, in the form of a
SingleCellExperiment. This represents a total of 229,047
cells x 51 channels.
spe same single cell data as for sce, but in the
SpatialExperiment format.
All data are downloaded from ExperimentHub and cached for local re-use.
Mapping between the three data objects is performed via variables located in
their metadata columns: mcols() for the CytoImageList
objects and ColData() for the SingleCellExperiment and
SpatialExperiment objects. Mapping at the image level can be
performed with the image_name or image_number variables.
Mapping between cell segmentation masks and single cell data is performed
with the cell_number variable, the values of which correspond to the
intensity values of the masks object. For practical examples, please
refer to the "Accessing IMC datasets" vignette.
This dataset was obtained as following (the names of the experimental
variables, located in the colData of the
SingleCellExperiment and SpatialExperiment
objects, are indicated in parentheses): i) Cells from four different
cell lines (cell_line) were seeded at three different densities
(treatment_concentration, relative densities) and grown for either 72
or 96 hours (treatment_time_point, duration in hours). In the
appropriate experimental conditions (see the paper for details), the cells
aggregate into 3D spheroids. ii) Cells were harvested and pooled into
60-well barcoding plates. iii) A pellet of each spheroid pool was
generated and cut into several 6 um-thick sections. iv) A subset of
these sections (site_id) were stained with an IMC panel and acquired
as one or more acquisitions (acquisition_id) containing multiple
spheres each. v) Spheres in these acquisitions were identified by
computer vision and cropped into individual images (image_number).
Other relevant cell metadata include:
treatment_name: experimental conditions in the format:
"Cell line name"_c"seeding density"_tp"time point".
cell_x/cell_y: cell centroid position in the image.
cell_area: area of the cell (um^2).
distance_rim: estimated distance to spheroid border.
distance_sphere: distance to spheroid section border.
distance_other_sphere: distance to the closest of the other
spheroid sections in the same image (if there is any).
distance_background: distance to background pixels.
For a full description of the other experimental variables, please refer to the publication (https://doi.org/10.15252/msb.20209798) and to the original dataset repository (https://doi.org/10.5281/zenodo.4271910).
The marker-associated metadata, including antibody information and metal
tags are stored in the rowData of the
SingleCellExperiment and SpatialExperiment
objects. The channels with names starting with "BC_" are the channels used
for barcoding. Post-transcriptional modification of the protein targets are
indicated in brackets.
The assay slots of the SingleCellExperiment and
SpatialExperiment objects contain three assays:
counts contains raw mean ion counts per cell.
exprs contains arsinh-transformed counts, with cofactor 1.
quant_norm contains counts censored at the 99th percentile
and scaled 0-1.
In addition, the altExp slot of the
SingleCellExperiment object contains another
SingleCellExperiment object where the counts matrix represents
raw mean ion counts for cells neighboring the current cell.
Neighborhood information, defined here as cells that are localized next to
each other, is stored as a SelfHits object in the colPairs
slot of the SingleCellExperiment and SpatialExperiment
objects. Cells in the SelfHits
object are represented by unique integers that map to the
cell_number_absolute column of colData(sce).
Dataset versions: a version argument can be passed to the function to
specify which dataset version should be retrieved.
`v0`: original version (Bioconductor <= 3.15).
`v1`: consistent object formatting across datasets.
File sizes:
`images`: size in memory = 21.2 Gb, size on disk = 860 Mb.
`masks`: size in memory = 426 Mb, size on disk = 12 Mb.
`sce`: size in memory = 564 Mb, size on disk = 319 Mb.
`spe`: size in memory = 596 Mb, size on disk = 320 Mb.
When storing images on disk, these need to be first fully read into memory before writing them to disk. This means the process of downloading the data is slower than directly keeping them in memory. However, downstream analysis will lose its memory overhead when storing images on disk.
Original source: Zanotelli et al. (2020): https://doi.org/10.15252/msb.20209798
Original link to raw data, also containing the entire dataset: https://doi.org/10.5281/zenodo.4271910
A SingleCellExperiment object with single cell data, a SpatialExperiment object with single cell data, a CytoImageList object containing multichannel images, or a CytoImageList object containing cell segmentation masks.
Nicolas Damond
Zanotelli VRT et al. (2020). A quantitative analysis of the interplay of environment, neighborhood, and cell state in 3D spheroids Mol Syst Biol 16(12), e9798.
# Load single cell data sce <- Zanotelli_2020_Spheroids(data_type = "sce") print(sce) # Display metadata Zanotelli_2020_Spheroids(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- Zanotelli_2020_Spheroids(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))# Load single cell data sce <- Zanotelli_2020_Spheroids(data_type = "sce") print(sce) # Display metadata Zanotelli_2020_Spheroids(data_type = "sce", metadata = TRUE) # Load masks on disk library(HDF5Array) masks <- Zanotelli_2020_Spheroids(data_type = "masks", on_disk = TRUE, h5FilesPath = getHDF5DumpDir()) print(head(masks))