1 Introduction

This vignette gives an introduction to handling and analyzing imaging mass cytometry (IMC) and other highly-multiplexed imaging data in R. The imcRtools package relies on expression and morphological features extracted from multi-channel images using corresponding segmentation masks. A description of data types and segmentation approaches can be found below (data types, segmentation). However, due to shared data structures, the functionalities of the imcRtools package are applicable to most highly multiplexed imaging modalities.

1.1 Overview

The imcRtools package exports functions and example data to perform the following analyses:

  1. Read in pre-processed data
  2. Perform spillover correction for IMC data
  3. Build and visualize spatial graphs
  4. Aggregate across neighbouring cells for spatial clustering
  5. Detect spatial patches of similar cell-types
  6. Test the attraction or avoidance between celltypes

To highlight the usability of these function, the imcRtools package also exports a number of test data files.

1.2 Highly multiplexed imaging

Highly multiplexed imaging techniques (Bodenmiller 2016) such as imaging mass cytometry (IMC) (Giesen et al. 2014), multiplexed ion beam imaging (MIBI) (Angelo et al. 2014) and cyclic immunofluorescence techniques (Lin et al. 2018, @Gut2018) acquire read-outs of the expression of tens of protein in a spatially resolved manner. After technology-dependent data pre-processing, the raw data files are comparable: multi-channel images for the dimensions x, y, and c, where x and y define the number of pixels (x * y) per image and c the number of proteins (also refered to as “markers”) measured per image. More info on the data types and further pre-processing can be found below.

Increased multiplexity compared to epitope-based techniques is achieved using single-cell resolved spatial transcriptomics techniques including MERFISH (Chen et al. 2015) and seqFISH (Lubeck et al. 2014). However, data produced by these techniques required different pre-processing steps. Nevertheless, analysis performed on the single-cell level is equally applicable.

1.2.1 Imaging mass cytometry

IMC (Giesen et al. 2014) is a highly multiplexed imaging approach to measure spatial protein abundance. In IMC, tissue sections on slides are stained with a mix of around 40 metal-conjugated antibodies prior to laser ablation with \(1\mu{}m\) resolution. The ablated material is transferred to a mass cytometer for time-of-flight detection of the metal ions (Giesen et al. 2014). At an ablation frequency of 200Hz, a 1mm x 1mm region can be acquired within 1.5 hours.

1.2.2 Data types

In the case of IMC, the raw data output are .mcd files containing all acquired regions per slide. During image pre-prcoessing these files are converted into individual multi-channel .tiff and OME-TIFF files. These file formats are supported by most open-source and commercial image analysis software, such as Fiji, QuPath or napari.

In R, these .tiff files can be read into individual Image objects or combined into a CytoImageList object supported by the cytomapper package.

1.2.3 Segmentation and feature extraction

The pixel resolution of most highly multiplexed imaging technologies including IMC support the detection of cellular structures. Therefore, a common processing step using multi-channel images is object segmentation. In this vignette, objects are defined as cells; however, also larger scale structures could be segmented.

There are multiple different image segmentation approaches available. However, imcRtools only supports direct reader functions for two segmentation strategies developed for highly multiplexed imaging technologies:

  1. The ImcSegmentationPipeline has been developed to give the user flexibility in how to perform channel selection and image segmentation. The underlying principle is to train a pixel classifier (using ilastik) on a number of selected channels to perform probabilistic classification of each pixel into: background, cytoplasm and nucleus. Based on these classification probabilities a CellProfiler pipeline performs cell segmentation and feature extraction.

  2. A containerized version of this pipeline is implemented in steinbock. steinbock further supports segmentation via the use of Mesmer from the DeepCell library (Greenwald et al. 2021).

While the output of both approaches is structured differently, the exported features are comparable:

  1. per cell: channel intensity, morphology and location
  2. cell-cell interactions exported as graph

By combining these with availabel channel information, the data can be read into a SpatialExperiment or SingleCellExperiment object (see below).

2 Example data

The imcRtools package contains a number of example data generated by the Hyperion imaging system for different purposes. The following section gives an overview of these files.

2.1 For spillover correction

To highlight the use of the imcRtools package for spillover correction, we provide four .txt files containing pixel intensities of four spotted metals.

Please refer to the spillover correction section for full details.

These files are accessible via:

path <- system.file("extdata/spillover", package = "imcRtools")

list.files(path, recursive = TRUE)
## [1] "Dy161.txt" "Dy162.txt" "Dy163.txt" "Dy164.txt"

2.2 Raw data in form of .txt files

IMC generates .mcd files storing the raw data or all acquired regions of interest (ROI). In addition, the raw pixel values are also stored in individual .txt files per ROI.

To highlight reading in raw data in form of .txt files, the imcRtools contains 3 sample acquisitions:

txt_files <- list.files(system.file("extdata/mockData/raw", 
                                    package = "imcRtools"))
txt_files
## [1] "20210305_NE_mockData2_ROI_001_1.txt" "20210305_NE_mockData2_ROI_002_2.txt"
## [3] "20210305_NE_mockData2_ROI_003_3.txt"

2.3 ImcSegmentationPipeline output data

IMC data preprocessing and segmentation can be performed using the ImcSegmentationPipeline. It generates a number of .csv files containing object/cell-specific and image-specific metadata.

The imcRtools package exports the read_cpout function as convenient reader function for outputs generated by the ImcSegmentationPipeline. For demonstration purposes, imcRtools contains the output of running the pipeline on a small example dataset:

path <- system.file("extdata/mockData/cpout", package = "imcRtools")

list.files(path, recursive = TRUE)
## [1] "Experiment.csv"           "Image.csv"               
## [3] "Object_relationships.csv" "cell.csv"                
## [5] "panel.csv"                "var_Image.csv"           
## [7] "var_cell.csv"

2.4 steinbock output data

The steinbock pipeline can be used to process, segment and extract features from IMC data. For more information, please refer to its documentation.

To highlight the functionality of imcRtools to read in single-cell data generated by steinbock, we provide a small toy dataset available at:

path <- system.file("extdata/mockData/steinbock", package = "imcRtools")

list.files(path, recursive = TRUE)
##  [1] "cells.csv"                              
##  [2] "images.csv"                             
##  [3] "intensities/20210305_NE_mockData1_1.csv"
##  [4] "intensities/20210305_NE_mockData1_2.csv"
##  [5] "intensities/20210305_NE_mockData1_3.csv"
##  [6] "intensities/20210305_NE_mockData2_1.csv"
##  [7] "intensities/20210305_NE_mockData2_2.csv"
##  [8] "intensities/20210305_NE_mockData2_3.csv"
##  [9] "intensities/20210305_NE_mockData3_1.csv"
## [10] "intensities/20210305_NE_mockData3_2.csv"
## [11] "intensities/20210305_NE_mockData3_3.csv"
## [12] "intensities/20210305_NE_mockData4_1.csv"
## [13] "intensities/20210305_NE_mockData4_2.csv"
## [14] "intensities/20210305_NE_mockData4_3.csv"
## [15] "intensities/20210305_NE_mockData5_1.csv"
## [16] "intensities/20210305_NE_mockData5_2.csv"
## [17] "intensities/20210305_NE_mockData5_3.csv"
## [18] "neighbors/20210305_NE_mockData1_1.csv"  
## [19] "neighbors/20210305_NE_mockData1_2.csv"  
## [20] "neighbors/20210305_NE_mockData1_3.csv"  
## [21] "neighbors/20210305_NE_mockData2_1.csv"  
## [22] "neighbors/20210305_NE_mockData2_2.csv"  
## [23] "neighbors/20210305_NE_mockData2_3.csv"  
## [24] "neighbors/20210305_NE_mockData3_1.csv"  
## [25] "neighbors/20210305_NE_mockData3_2.csv"  
## [26] "neighbors/20210305_NE_mockData3_3.csv"  
## [27] "neighbors/20210305_NE_mockData4_1.csv"  
## [28] "neighbors/20210305_NE_mockData4_2.csv"  
## [29] "neighbors/20210305_NE_mockData4_3.csv"  
## [30] "neighbors/20210305_NE_mockData5_1.csv"  
## [31] "neighbors/20210305_NE_mockData5_2.csv"  
## [32] "neighbors/20210305_NE_mockData5_3.csv"  
## [33] "panel.csv"                              
## [34] "regionprops/20210305_NE_mockData1_1.csv"
## [35] "regionprops/20210305_NE_mockData1_2.csv"
## [36] "regionprops/20210305_NE_mockData1_3.csv"
## [37] "regionprops/20210305_NE_mockData2_1.csv"
## [38] "regionprops/20210305_NE_mockData2_2.csv"
## [39] "regionprops/20210305_NE_mockData2_3.csv"
## [40] "regionprops/20210305_NE_mockData3_1.csv"
## [41] "regionprops/20210305_NE_mockData3_2.csv"
## [42] "regionprops/20210305_NE_mockData3_3.csv"
## [43] "regionprops/20210305_NE_mockData4_1.csv"
## [44] "regionprops/20210305_NE_mockData4_2.csv"
## [45] "regionprops/20210305_NE_mockData4_3.csv"
## [46] "regionprops/20210305_NE_mockData5_1.csv"
## [47] "regionprops/20210305_NE_mockData5_2.csv"
## [48] "regionprops/20210305_NE_mockData5_3.csv"
## [49] "steinbock.sh"

The example data related to the ImcSegmentationPipeline and steinbock can also be accessed online.

3 Read in IMC data

The imcRtools package supports reading in data generated by the ImcSegmentationPipeline or steinbock pipeline.

To read in the outpout data into a SpatialExperiment or SingleCellExperiment, the imcRtools package exports the read_cpout function.

By default, the single-cell data is read into a SpatialExperiment object. Here, the extracted channel- and cell-specific intensities are stored in the counts(spe) slot. All morphological features are stored in colData(spe) and the spatial locations of the cells are stored in spatialCoords(spe). The interaction graph is stored in colPair(spe, "neighbourhood").

Alternatively, the data can be read into a SingleCellExperiment object. The only difference is the lack of spatialCoords(sce). Here, the spatial coordinates are stored in colData(spe)$Pos_X and colData(spe)$Pos_Y.

3.1 Read in CellProfiler output

The ImcSegmentationPipeline produces a number of output files. By default, all single-cell features are measured and exported. To browse and select the features of interest, the imcRtools package provides the show_cpout_features function:

path <- system.file("extdata/mockData/cpout", package = "imcRtools")

show_cpout_features(path)

By default, read_cpout will read in the mean intensity per channel and cell from “hot pixel” filtered image stacks specified via intensities = "Intensity_MeanIntensity_FullStackFiltered". Please refer to ?read_cpout for the full documentation.

cur_path <- system.file("extdata/mockData/cpout", package = "imcRtools")

# Read as SpatialExperiment
(spe <- read_cpout(cur_path, graph_file = "Object_relationships.csv"))
## class: SpatialExperiment 
## dim: 5 219 
## metadata(0):
## assays(1): counts
## rownames(5): Ag107 Pr141 Sm147 Eu153 Yb172
## rowData names(5): Tube.Number Metal.Tag Target ilastik full
## colnames: NULL
## colData names(6): sample_id ObjectNumber ... Metadata_acid
##   Metadata_description
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):
## spatialData names(0) :
## spatialCoords names(2) : Pos_X Pos_Y
## imgData names(1): sample_id
# Read as SingleCellExperiment
(sce <- read_cpout(cur_path, graph_file = "Object_relationships.csv", 
                   return_as = "sce"))
## class: SingleCellExperiment 
## dim: 5 219 
## metadata(0):
## assays(1): counts
## rownames(5): Ag107 Pr141 Sm147 Eu153 Yb172
## rowData names(5): Tube.Number Metal.Tag Target ilastik full
## colnames: NULL
## colData names(8): sample_id ObjectNumber ... Metadata_acid
##   Metadata_description
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):

3.2 Read in steinbock output

Single-cell data and all associated metadata (e.g. spatial location, morphology and interaction graphs) as produced by the steinbock pipeline can be read in using the read_steinbock function:

cur_path <- system.file("extdata/mockData/steinbock", package = "imcRtools")

# Read as SpatialExperiment
(spe <- read_steinbock(cur_path))
## class: SpatialExperiment 
## dim: 5 218 
## metadata(0):
## assays(1): counts
## rownames(5): Ag107 Cytokeratin 5 Laminin YBX1 H3K27Ac
## rowData names(6): channel name ... deepcell Tube.Number
## colnames: NULL
## colData names(8): sample_id ObjectNumber ... width_px height_px
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):
## spatialData names(0) :
## spatialCoords names(2) : Pos_X Pos_Y
## imgData names(1): sample_id
# Read as SingleCellExperiment
(sce <- read_steinbock(cur_path, return_as = "sce"))
## class: SingleCellExperiment 
## dim: 5 218 
## metadata(0):
## assays(1): counts
## rownames(5): Ag107 Cytokeratin 5 Laminin YBX1 H3K27Ac
## rowData names(6): channel name ... deepcell Tube.Number
## colnames: NULL
## colData names(10): sample_id ObjectNumber ... width_px height_px
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):

For more information, please refer to ?read_steinbock.

3.3 Read raw .txt files into Image objects

For reading in and visualization of multi-channel images and segmentation masks, please refer to the cytomapper package. The imcRtools package however supports reading in raw .txt files generated by the Hyperion imaging system into a CytoImageList object; a data container exported by cytomapper.

The user needs to provide a path from which all .txt files will be read in:

path <- system.file("extdata/mockData/raw", package = "imcRtools")

cur_CytoImageList <- readImagefromTXT(path)
cur_CytoImageList
## CytoImageList containing 3 image(s)
## names(3): 20210305_NE_mockData2_ROI_001_1 20210305_NE_mockData2_ROI_002_2 20210305_NE_mockData2_ROI_003_3 
## Each image contains 5 channel(s)
## channelNames(5): Ag107Di Pr141Di Sm147Di Eu153Di Yb172Di

By specifying the pattern argument, individual or a subset of files can be read in. For more information, please refer to ?readImagefromTXT.

4 Spillover correction

When acquiring IMC images, pixel intensities can be influenced by spillover from neighboring channels. To correct for this, Chevrier et al. have developed a staining protocol to acquire individually spotted metal isotopes (Chevrier et al. 2017). Based on these measurements, spillover into neighboring channels can be quantified to correct pixel intensities.

The imcRtools package provides helper functions that facilitate the correction of spillover for IMC data. For a full tutorial, please refer to the IMC data analysis book.

4.1 Read in the single-spot acquisitions

In the first step, the pixel intensities of individually spotted metals need to be read into a SingleCellExperiment container for downstream use with the CATALYST package. For this, the readSCEfromTXT function can be used:

path <- system.file("extdata/spillover", package = "imcRtools")
sce <- readSCEfromTXT(path) 
## Spotted channels:  Dy161, Dy162, Dy163, Dy164
## Acquired channels:  Dy161, Dy162, Dy163, Dy164
## Channels spotted but not acquired:  
## Channels acquired but not spotted:
sce
## class: SingleCellExperiment 
## dim: 4 400 
## metadata(0):
## assays(1): counts
## rownames(4): 161Dy(Dy161Di) 162Dy(Dy162Di) 163Dy(Dy163Di)
##   164Dy(Dy164Di)
## rowData names(2): channel_name marker_name
## colnames(400): Dy161.1 Dy161.2 ... Dy164.99 Dy164.100
## colData names(9): Start_push End_push ... sample_metal sample_mass
## reducedDimNames(0):
## mainExpName: NULL
## altExpNames(0):

Here, the example metal spot files are read in. The spot information are stored in the colData(sce) slot and channel information are stored in rowData(sce). Each column represents a single pixel.

4.2 Quality control on single-spot acquisitions

In the next step, it is crucial to identify potentially mislabeled spots or spots with low pixel intensities. The imcRtools package exports the plotSpotHeatmap function, which visualizes the aggregated (default median) pixel intensities per spot and per metal:

plotSpotHeatmap(sce)

Here, high median pixel intensities can be observed in each spot and their corresponding channels (visualized on the log10 scale by default). To quickly identify spot/channel combinations with low signal, the threshold parameter can be set:

plotSpotHeatmap(sce, log = FALSE, threshold = 200)

4.3 Consecutive pixel binning

If pixel intensities are low, spillover estimation might not be robust. Therefore, the binAcrossPixels function can be used to sum consecutive pixels and enhance the acquired signal. This step is optional for spillover estimation.

sce2 <- binAcrossPixels(sce, bin_size = 5)

plotSpotHeatmap(sce2, log = FALSE, threshold = 200)

4.4 Pixel filtering

Prior to spillover estimation, the CATALYST package provides the assignPrelim, estCutoffs and applyCutoffs functions to estimate the spotted mass for each pixel based on their channel intensities. For more information on the spillover estimation and correction, please refer to the CATALYST vignette.

This estimation can be used to identify pixels that cannot be easily assigned to their spotted mass, potentially indicating pixels with weak signal. To remove these pixels, the filterPixels function can be used. This function further removes pixels assigned to masses, which only contain very few pixels.

library(CATALYST)

bc_key <- as.numeric(unique(sce$sample_mass))
assay(sce, "exprs") <- asinh(counts(sce)/5)
sce <- assignPrelim(sce, bc_key = bc_key)
sce <- estCutoffs(sce)
sce <- applyCutoffs(sce)

# Filter out mislabeled pixels
sce <- filterPixels(sce)

table(sce$bc_id, sce$sample_mass)
##      
##       161 162 163 164
##   0     0   0   0   2
##   161 100   0   0   0
##   162   0 100   0   0
##   163   0   0 100   0
##   164   0   0   0  98

4.5 Estimating the spillover matrix

Finally, the pre-processed SiingleCellExperiment object can be used to generate the spillover matrix using the CATALYST::computeSpillmat function:

sce <- computeSpillmat(sce)
metadata(sce)$spillover_matrix
##             Dy161Di     Dy162Di     Dy163Di     Dy164Di
## Dy161Di 1.000000000 0.031443129 0.009734712 0.006518048
## Dy162Di 0.015715159 1.000000000 0.048116187 0.008250039
## Dy163Di 0.003809504 0.012159704 1.000000000 0.020214651
## Dy164Di 0.005058069 0.008457546 0.028912343 1.000000000

This spillover matrix can be directly applied to compensate the summarized pixel intensities per cell and per channel as described here.

5 Spatial analysis

The following section will highlight functions for spatial analyses of the data.

5.1 Constructing graphs

When following the ImcSegmentationPipeline or steinbock and reading in the data using the corresponding functions, the generated graphs are automatically stored in the colPair(spe, "neighborhood") slot. Alternatively, the buildSpatialGraph function in the imcRtools package constructs interaction graphs using either (i) cell-centroid expansion, (ii) k-nearest neighbor search or (iii) delaunay triangulation.

library(cytomapper)
data("pancreasSCE")

pancreasSCE <- buildSpatialGraph(pancreasSCE, img_id = "ImageNb",
                                 type = "expansion",
                                 threshold = 20)
pancreasSCE <- buildSpatialGraph(pancreasSCE, img_id = "ImageNb",
                                 type = "knn",
                                 k = 5)
pancreasSCE <- buildSpatialGraph(pancreasSCE, img_id = "ImageNb",
                                 type = "delaunay")

colPairNames(pancreasSCE)
## [1] "expansion_interaction_graph" "knn_interaction_graph"      
## [3] "delaunay_interaction_graph"

When setting type = "knn", by default a directional graph will be build. Setting directed = FALSE will create bi-directional edges for each pair of cells that are connected by at least one edge in the directed setting.

5.2 Graph/cell visualization

The cells’ locations and constructed graphs can be visualized using the plotSpatial function. Here, cells are referred to as “nodes” and cell-cell interactions are referred to as “edges”. All visual attributes of the nodes and edges can be set. Either by specifying a variable in colData(spe), a marker name or a single entry using the *_fix parameters.

library(ggplot2)
library(ggraph)

plotSpatial(pancreasSCE,
            img_id = "ImageNb",
            node_color_by = "CellType",
            node_shape_by = "ImageNb",
            node_size_by = "Area",
            draw_edges = TRUE,
            colPairName = "knn_interaction_graph",
            directed = FALSE)

# Colored by expression and with arrows
plotSpatial(pancreasSCE,
            img_id = "ImageNb",
            node_color_by = "PIN",
            assay_type = "exprs",
            node_size_fix = 3,
            edge_width_fix = 0.2,
            draw_edges = TRUE,
            colPairName = "knn_interaction_graph",
            directed = TRUE,
            arrow = grid::arrow(length = grid::unit(0.1, "inch")),
            end_cap = ggraph::circle(0.05, "cm"))

# Subsetting the SingleCellExperiment
plotSpatial(pancreasSCE[,pancreasSCE$Pattern],
            img_id = "ImageNb",
            node_color_by = "CellType",
            node_size_fix = 1,
            draw_edges = TRUE,
            colPairName = "knn_interaction_graph",
            directed = TRUE,
            scales = "fixed")

The returned object can be further modified using the ggplot2 logic. This includes changing the node color, shape and size using scale_color_*, scale_shape_* and scale_size_*. Edge attributes can be altered using the scale_edge_* function exported by ggraph,

5.3 Neighborhood aggregation

The aggregateNeighbors function can be used to aggregate features of all neighboring cells for each individual cell. This function operates in two settings.

  1. metadata: when aggregating by cell-specific metadata, the function computes the relative frequencies of all entries to colData(sce)[[count_by]] within the direct neighborhood of each cell.
  2. expression: the expression counts of neighboring cells are aggregated using the specified statistic (defaults to mean).

Each cell’s neighborhood is defined as endpoints of edges stored in colPair(sce, colPairName).

pancreasSCE <- aggregateNeighbors(pancreasSCE,
                                  colPairName = "knn_interaction_graph",
                                  aggregate_by = "metadata",
                                  count_by = "CellType")
head(pancreasSCE$aggregatedNeighbors)
## DataFrame with 6 rows and 3 columns
##   celltype_A celltype_B celltype_C
##    <numeric>  <numeric>  <numeric>
## 1          0        0.0        1.0
## 2          0        0.2        0.8
## 3          0        0.0        1.0
## 4          0        0.0        1.0
## 5          0        0.0        1.0
## 6          0        0.0        1.0
pancreasSCE <- aggregateNeighbors(pancreasSCE,
                                  colPairName = "knn_interaction_graph",
                                  aggregate_by = "expression",
                                  assay_type =  "exprs")
head(pancreasSCE$mean_aggregatedExpression)
## DataFrame with 6 rows and 5 columns
##          H3      CD99       PIN      CD8a       CDH
##   <numeric> <numeric> <numeric> <numeric> <numeric>
## 1   2.32500  0.860329  0.092871  0.725000   2.51264
## 2   2.88022  1.629762  0.319527  0.207873   2.46486
## 3   3.10829  0.735389  0.190616  0.255515   1.89484
## 4   2.55842  0.773342  0.124545  0.188629   2.51084
## 5   2.44287  1.126240  0.252129  0.200261   2.61336
## 6   2.65059  0.903869  0.181792  0.196691   2.16434

The returned entries can now be used for clustering to group cells based on their environment (either by aggregated categorical features or expression).

cur_cluster <- kmeans(pancreasSCE$aggregatedNeighbors, centers = 3)
pancreasSCE$clustered_neighbors <- factor(cur_cluster$cluster)

plotSpatial(pancreasSCE,
            img_id = "ImageNb",
            node_color_by = "CellType",
            node_size_fix = 4,
            edge_width_fix = 2,
            edge_color_by = "clustered_neighbors",
            draw_edges = TRUE,
            colPairName = "knn_interaction_graph",
            directed = FALSE,
            nodes_first = FALSE) +
    scale_color_brewer(palette = "Set2") +
    scale_edge_color_brewer(palette = "Set1")

To exclude cells that are close to the image border, the imcRtools package exports the findBorderCells function.

pancreasSCE <- findBorderCells(pancreasSCE, 
                               img_id = "ImageNb", 
                               border_dist = 10)

plotSpatial(pancreasSCE[,!pancreasSCE$border_cells],
            img_id = "ImageNb",
            node_color_by = "CellType",
            node_size_fix = 4,
            edge_width_fix = 2,
            edge_color_by = "clustered_neighbors",
            draw_edges = TRUE,
            colPairName = "knn_interaction_graph",
            directed = FALSE,
            nodes_first = FALSE) +
    scale_color_brewer(palette = "Set2") +
    scale_edge_color_brewer(palette = "Set1")

Excluding border cells can be useful when incorrectly connected cells are observed at image borders.

5.4 Patch detection

An alternative and supervised way of detecting regions with similar types of cells is available via the patchDetection function. Here, the user defines which cells should be used for patch detection via the patch_cells parameter. A patch is defined as a set of cells as defined by patch_cells which are weakly connected in the graph in colPair(object, colPairname).

Below, the patchDetection function is demonstrated using the previously computed expansion graph and defining cells of celltype_B as the cells of interest. Here, the function additionally draws a concave hull around the detected patch, expands the hull by 20\(\mu{}m\) and defines all cells within this expanded hulls as patch cells.

pancreasSCE <- patchDetection(pancreasSCE, 
                              patch_cells = pancreasSCE$CellType == "celltype_B",
                              colPairName = "expansion_interaction_graph",
                              expand_by = 20, 
                              img_id = "ImageNb")
 
plotSpatial(pancreasSCE, img_id = "ImageNb", node_color_by = "patch_id")

Patches that only consist of 1 or 2 cells cannot be expanded.

5.5 Neighborhood permutation testing

The following section describes how to observe and test the average number of interactions between cell labels (e.g. cell-types) within grouping levels (e.g. images). For full descriptions of the testing approaches, please refer to Shapiro et al., Nature Methods (Schapiro et al. 2017) and Schulz et al., Cell Systems (Schulz et al. 2018)

The imcRtools package exports the countInteractions and testInteractions functions, which summarize all cell-cell interactions per grouping level (e.g. image). As a result, a table is returned where each row represents one of all possible cell-type/cell-type interactions among all grouping levels. Missing entries or NAs indicate missing cell-type labels for this grouping level. The next section gives details on how interactions are summarized.

5.5.1 Summarizing interactions

The countInteractions function counts the number of edges (interactions) between each set of unique cell labels per grouping level. Simplified, it counts for each cell of type A the number of neighbors of type B. This count is averaged within each unique grouping level (e.g. image) in three different ways:

  1. method = "classic": The count is divided by the total number of cells of type A. The final count can be interpreted as “How many neighbors of type B does a cell of type A have on average?”

  2. method = "histocat": The count is divided by the number of cells of type A that have at least one neighbor of type B. The final count can be interpreted as “How many many neighbors of type B has a cell of type A on average, given it has at least one neighbor of type B?”

  3. method = "patch": For each cell, the count is binarized to 0 (less than patch_size neighbors of type B) or 1 (more or equal to patch_size neighbors of type B). The binarized counts are averaged across all cells of type A. The final count can be interpreted as “What fraction of cells of type A have at least a given number of neighbors of type B?”

The countInteractions returns a DataFrame containing the summarized counts (ct) for all combinations of from_label, to_label and group_by.

out <- countInteractions(pancreasSCE,
                         group_by = "ImageNb",
                         label = "CellType",
                         method = "classic",
                         colPairName = "knn_interaction_graph")
out
## DataFrame with 27 rows and 4 columns
##      group_by from_label   to_label        ct
##     <integer>   <factor>   <factor> <numeric>
## 1           1 celltype_A celltype_A  2.823529
## 2           1 celltype_A celltype_B  0.823529
## 3           1 celltype_A celltype_C  1.352941
## 4           1 celltype_B celltype_A  2.000000
## 5           1 celltype_B celltype_B  0.625000
## ...       ...        ...        ...       ...
## 23          3 celltype_B celltype_B   4.00000
## 24          3 celltype_B celltype_C   1.00000
## 25          3 celltype_C celltype_A        NA
## 26          3 celltype_C celltype_B   1.13115
## 27          3 celltype_C celltype_C   3.86885

5.5.2 Testing for significance

In the next instance, one can test if the obtained count is larger or smaller compared to what is expected from a random distribution of cell labels. For this, the testInteractions function permutes the cell labels iter times and counts interactions as described above. This approach generates a distribution of the interaction count under a random distribution of cell labels. The observed interaction count is compared against this Null distribution to derive empirical p-values:

p_gt: fraction of perturbations equal or greater than the observed count

p_lt: fraction of perturbations equal or less than the observed count

Based on these empirical p-values, the interaction score (attraction or avoidance), overall p value and significance by comparison to p_treshold (sig and sigval) are derived. All results are returned in form of a DataFrame.

out <- testInteractions(pancreasSCE,
                        group_by = "ImageNb",
                        label = "CellType",
                        method = "classic",
                        colPairName = "knn_interaction_graph")
out
## DataFrame with 27 rows and 10 columns
##      group_by from_label   to_label        ct        p_gt        p_lt
##     <integer>   <factor>   <factor> <numeric>   <numeric>   <numeric>
## 1           1 celltype_A celltype_A  2.823529 0.000999001 1.000000000
## 2           1 celltype_A celltype_B  0.823529 0.000999001 1.000000000
## 3           1 celltype_A celltype_C  1.352941 1.000000000 0.000999001
## 4           1 celltype_B celltype_A  2.000000 0.000999001 1.000000000
## 5           1 celltype_B celltype_B  0.625000 0.132867133 0.921078921
## ...       ...        ...        ...       ...         ...         ...
## 23          3 celltype_B celltype_B   4.00000 0.000999001 1.000000000
## 24          3 celltype_B celltype_C   1.00000 1.000000000 0.000999001
## 25          3 celltype_C celltype_A        NA          NA          NA
## 26          3 celltype_C celltype_B   1.13115 1.000000000 0.000999001
## 27          3 celltype_C celltype_C   3.86885 0.000999001 1.000000000
##     interaction           p       sig    sigval
##       <logical>   <numeric> <logical> <numeric>
## 1          TRUE 0.000999001      TRUE         1
## 2          TRUE 0.000999001      TRUE         1
## 3         FALSE 0.000999001      TRUE        -1
## 4          TRUE 0.000999001      TRUE         1
## 5          TRUE 0.132867133     FALSE         0
## ...         ...         ...       ...       ...
## 23         TRUE 0.000999001      TRUE         1
## 24        FALSE 0.000999001      TRUE        -1
## 25           NA          NA        NA        NA
## 26        FALSE 0.000999001      TRUE        -1
## 27         TRUE 0.000999001      TRUE         1

6 Contributions

Large chunks of the code of the imcRtools package is based on the work of Vito Zanotelli. Vito has co-developed the spillover correction approach and translated the interaction testing approach developed by Denis Shapiro and Jana Fischer into R (formerly known as the neighbouRhood R package). Jana has furthermore added the “patch” method for interaction counting and testing. Tobias Hoch has written the first version of reading in the ImcSegmentationPipeline output and the patchDetection function. Daniel Schulz has build the aggregateNeighbors function and contributed to developing the spatial clustering approach.

Session info

## R version 4.1.2 (2021-11-01)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows Server 2012 R2 x64 (build 9600)
## 
## Matrix products: default
## 
## locale:
## [1] LC_COLLATE=C                          
## [2] LC_CTYPE=English_United States.1252   
## [3] LC_MONETARY=English_United States.1252
## [4] LC_NUMERIC=C                          
## [5] LC_TIME=English_United States.1252    
## 
## attached base packages:
## [1] stats4    stats     graphics  grDevices utils     datasets  methods  
## [8] base     
## 
## other attached packages:
##  [1] ggraph_2.0.5                ggplot2_3.3.5              
##  [3] cytomapper_1.6.0            EBImage_4.36.0             
##  [5] CATALYST_1.18.0             imcRtools_1.0.1            
##  [7] SpatialExperiment_1.4.0     SingleCellExperiment_1.16.0
##  [9] SummarizedExperiment_1.24.0 Biobase_2.54.0             
## [11] GenomicRanges_1.46.1        GenomeInfoDb_1.30.0        
## [13] IRanges_2.28.0              S4Vectors_0.32.3           
## [15] BiocGenerics_0.40.0         MatrixGenerics_1.6.0       
## [17] matrixStats_0.61.0          BiocStyle_2.22.0           
## 
## loaded via a namespace (and not attached):
##   [1] scattermore_0.7             flowWorkspace_4.6.0        
##   [3] R.methodsS3_1.8.1           tidyr_1.1.4                
##   [5] bit64_4.0.5                 knitr_1.36                 
##   [7] irlba_2.3.3                 multcomp_1.4-17            
##   [9] DelayedArray_0.20.0         R.utils_2.11.0             
##  [11] data.table_1.14.2           RCurl_1.98-1.5             
##  [13] doParallel_1.0.16           generics_0.1.1             
##  [15] flowCore_2.6.0              ScaledMatrix_1.2.0         
##  [17] TH.data_1.1-0               terra_1.4-22               
##  [19] cowplot_1.1.1               proxy_0.4-26               
##  [21] ggpointdensity_0.1.0        bit_4.0.4                  
##  [23] tzdb_0.2.0                  xml2_1.3.2                 
##  [25] httpuv_1.6.3                assertthat_0.2.1           
##  [27] viridis_0.6.2               xfun_0.28                  
##  [29] hms_1.1.1                   jquerylib_0.1.4            
##  [31] evaluate_0.14               promises_1.2.0.1           
##  [33] fansi_0.5.0                 Rgraphviz_2.38.0           
##  [35] igraph_1.2.8                DBI_1.1.1                  
##  [37] htmlwidgets_1.5.4           purrr_0.3.4                
##  [39] ellipsis_0.3.2              crosstalk_1.2.0            
##  [41] dplyr_1.0.7                 ggcyto_1.22.0              
##  [43] ggnewscale_0.4.5            ggpubr_0.4.0               
##  [45] backports_1.4.0             V8_3.6.0                   
##  [47] cytolib_2.6.0               bookdown_0.24              
##  [49] svgPanZoom_0.3.4            RcppParallel_5.1.4         
##  [51] sparseMatrixStats_1.6.0     vctrs_0.3.8                
##  [53] abind_1.4-5                 withr_2.4.2                
##  [55] ggforce_0.3.3               aws.signature_0.6.0        
##  [57] vroom_1.5.6                 svglite_2.0.0              
##  [59] cluster_2.1.2               crayon_1.4.2               
##  [61] drc_3.0-1                   labeling_0.4.2             
##  [63] edgeR_3.36.0                pkgconfig_2.0.3            
##  [65] units_0.7-2                 tweenr_1.0.2               
##  [67] vipor_0.4.5                 rlang_0.4.12               
##  [69] lifecycle_1.0.1             sandwich_3.0-1             
##  [71] rsvd_1.0.5                  polyclip_1.10-0            
##  [73] graph_1.72.0                tiff_0.1-10                
##  [75] Matrix_1.3-4                raster_3.5-2               
##  [77] carData_3.0-4               Rhdf5lib_1.16.0            
##  [79] zoo_1.8-9                   base64enc_0.1-3            
##  [81] beeswarm_0.4.0              RTriangle_1.6-0.10         
##  [83] ggridges_0.5.3              GlobalOptions_0.1.2        
##  [85] pheatmap_1.0.12             png_0.1-7                  
##  [87] viridisLite_0.4.0           rjson_0.2.20               
##  [89] bitops_1.0-7                shinydashboard_0.7.2       
##  [91] R.oo_1.24.0                 ConsensusClusterPlus_1.58.0
##  [93] KernSmooth_2.23-20          rhdf5filters_1.6.0         
##  [95] DelayedMatrixStats_1.16.0   shape_1.4.6                
##  [97] classInt_0.4-3              stringr_1.4.0              
##  [99] readr_2.1.0                 jpeg_0.1-9                 
## [101] rstatix_0.7.0               ggsignif_0.6.3             
## [103] aws.s3_0.3.21               beachmat_2.10.0            
## [105] scales_1.1.1                magrittr_2.0.1             
## [107] plyr_1.8.6                  hexbin_1.28.2              
## [109] zlibbioc_1.40.0             compiler_4.1.2             
## [111] dqrng_0.3.0                 concaveman_1.1.0           
## [113] RColorBrewer_1.1-2          plotrix_3.8-2              
## [115] clue_0.3-60                 XVector_0.34.0             
## [117] ncdfFlow_2.40.0             FlowSOM_2.2.0              
## [119] MASS_7.3-54                 tidyselect_1.1.1           
## [121] stringi_1.7.5               RProtoBufLib_2.6.0         
## [123] highr_0.9                   yaml_2.2.1                 
## [125] BiocSingular_1.10.0         locfit_1.5-9.4             
## [127] latticeExtra_0.6-29         ggrepel_0.9.1              
## [129] grid_4.1.2                  sass_0.4.0                 
## [131] tools_4.1.2                 parallel_4.1.2             
## [133] CytoML_2.6.0                circlize_0.4.13            
## [135] foreach_1.5.1               gridExtra_2.3              
## [137] farver_2.1.0                Rtsne_0.15                 
## [139] DropletUtils_1.14.1         digest_0.6.28              
## [141] BiocManager_1.30.16         shiny_1.7.1                
## [143] Rcpp_1.0.7                  car_3.0-12                 
## [145] broom_0.7.10                scuttle_1.4.0              
## [147] later_1.3.0                 httr_1.4.2                 
## [149] sf_1.0-4                    ComplexHeatmap_2.10.0      
## [151] colorspace_2.0-2            XML_3.99-0.8               
## [153] splines_4.1.2               RBGL_1.70.0                
## [155] scater_1.22.0               graphlayouts_0.7.2         
## [157] sp_1.4-6                    systemfonts_1.0.3          
## [159] xtable_1.8-4                jsonlite_1.7.2             
## [161] tidygraph_1.2.0             R6_2.5.1                   
## [163] pillar_1.6.4                htmltools_0.5.2            
## [165] mime_0.12                   nnls_1.4                   
## [167] glue_1.5.0                  fastmap_1.1.0              
## [169] DT_0.20                     BiocParallel_1.28.2        
## [171] BiocNeighbors_1.12.0        fftwtools_0.9-11           
## [173] class_7.3-19                codetools_0.2-18           
## [175] mvtnorm_1.1-3               utf8_1.2.2                 
## [177] lattice_0.20-45             bslib_0.3.1                
## [179] tibble_3.1.6                curl_4.3.2                 
## [181] ggbeeswarm_0.6.0            colorRamps_2.3             
## [183] gtools_3.9.2                magick_2.7.3               
## [185] survival_3.2-13             limma_3.50.0               
## [187] rmarkdown_2.11              munsell_0.5.0              
## [189] e1071_1.7-9                 GetoptLong_1.0.5           
## [191] rhdf5_2.38.0                GenomeInfoDbData_1.2.7     
## [193] iterators_1.0.13            HDF5Array_1.22.1           
## [195] reshape2_1.4.4              gtable_0.3.0

References

Angelo, Michael, Sean C. Bendall, Rachel Finck, Matthew B. Hale, Chuck Hitzman, Alexander D. Borowsky, Richard M. Levenson, et al. 2014. “Multiplexed Ion Beam Imaging of Human Breast Tumors.” Nature Medicine 20 (4): 436–42.

Bodenmiller, Bernd. 2016. “Multiplexed Epitope-Based Tissue Imaging for Discovery and Healthcare Applications.” Cell Systems 2: 225–38.

Chen, Kok Hao, Alistair N. Boettiger, Jeffrey R. Moffitt, Siyuan Wang, and Xiaowei Zhuang. 2015. “Spatially Resolved, Highly Multiplexed Rna Profiling in Single Cells.” Science 348: aaa6090.

Chevrier, Stéphane, Helena L. Crowell, Vito R. T. Zanotelli, Stefanie Engler, Mark D. Robinson, and Bernd Bodenmiller. 2017. “Compensation of Signal Spillover in Suspension and Imaging Mass Cytometry.” Cell Systems 6: 612–20.

Giesen, Charlotte, Hao A. O. Wang, Denis Schapiro, Nevena Zivanovic, Andrea Jacobs, Bodo Hattendorf, Peter J. Schüffler, et al. 2014. “Highly Multiplexed Imaging of Tumor Tissues with Subcellular Resolution by Mass Cytometry.” Nature Methods 11 (4): 417–22.

Greenwald, Noah F, Geneva Miller, Erick Moen, Alex Kong, Adam Kagel, Christine Camacho, Brianna J Mcintosh, et al. 2021. “Whole-Cell Segmentation of Tissue Images with Human-Level Performance Using Large-Scale Data Annotation and Deep Learning.” bioRxiv, 1–29.

Gut, Gabriele, Markus D Herrmann, and Lucas Pelkmans. 2018. “Multiplexed Protein Maps Link Subcellular Organization to Cellular States.” Science 361: 1–13.

Lin, Jia-Ren, Benjamin Izar, Shu Wang, Clarence Yapp, Shaolin Mei, Parin M. Shah, Sandro Santagata, and Peter K. Sorger. 2018. “Highly Multiplexed Immunofluorescence Imaging of Human Tissues and Tumors Using T-Cycif and Conventional Optical Microscopes.” eLife 7: 1–46.

Lubeck, Eric, Ahmet F Coskun, Timur Zhiyentayev, Mubhij Ahmad, and Long Cai. 2014. “Single-Cell in Situ Rna Profiling by Sequential Hybridization.” Nature Methods 11: 360–61.

Schapiro, Denis, Hartland W Jackson, Swetha Raghuraman, Jana R Fischer, Vito RT Zanotelli, Daniel Schulz, Charlotte Giesen, Raúl Catena, Zsuzsanna Varga, and Bernd Bodenmiller. 2017. “HistoCAT: Analysis of Cell Phenotypes and Interactions in Multiplex Image Cytometry Data.” Nature Methods 14: 873–76.

Schulz, Daniel, Vito RT Zanotelli, Rana R Fischer, Denis Schapiro, Stefanie Engler, Xiao-Kang Lun, Hartland W Jackson, and Bernd Bodenmiller. 2018. “Simultaneous Multiplexed Imaging of mRNA and Proteins with Subcellular Resolution in Breast Cancer Tissue Samples by Mass Cytometry.” Cell Systems 6: 25–36.e5.