Package 'scider'

Annotate all cells with contour level of cell type-specific density.

Description

Annotate all cells with contour level of cell type-specific density.

Usage

allocateCells(spe, to.roi = TRUE, to.contour = TRUE)

Arguments

spe	A SpatialExperiment object.
to.roi	Logical. Whether to allocate cells to ROIs.
to.contour	Logical. Whether to allocate cells to contour levels.

Value

A SpatialExperiment object. An extra column is added to the colData.

Examples

data("xenium_bc_spe")

spe <- gridDensity(spe)

coi <- "Breast cancer"

spe <- findROI(spe, coi = coi)

spe <- allocateCells(spe)

---

Check which cells are in which regions

Description

Check which cells are in which regions

Usage

cellsInRegion(spe, region, name_to, NA_level = "0", levels = NULL)

Arguments

spe	A SpatialExperiment object.
region	List or an sf object that represents a region or an ROI.
name_to	Colname in colData(spe) to store the annotation.
NA_level	Label for cells not falling in any of the regions. Default to 0.
levels	Factor levels.

Value

A SpatialExperiment object. The region information of each cell is stored in the colData.

---

Perform kernel density estimation on SpatialExperiment

Description

Perform kernel density estimation on SpatialExperiment

Usage

computeDensity(
  spe,
  mode = "pixels",
  kernel = "gaussian",
  bandwidth = NULL,
  weights = NULL,
  ngrid.x = 100,
  ngrid.y = NULL,
  grid.length.x = NULL,
  grid.length.y = NULL,
  xlim = NULL,
  ylim = NULL,
  diggle = FALSE
)

Arguments

spe	A SpatialExperiment object.
mode	Choose either points or pixels. Specifying whether to compute the density at a grid pixel location or at at point.
kernel	The smoothing kernel. Options are gaussian, epanechnikov, quartic or disc.
bandwidth	The smoothing bandwidth. By default performing automatic bandwidth selection using cross-validation using function spatstat.explore::bw.diggle.
weights	Optional weights to be attached to the points.
ngrid.x	Number of grids in the x-direction. Default to 100.
ngrid.y	Number of grids in the y-direction.
grid.length.x	Grid length in the x-direction.
grid.length.y	Grid length in the y-direction.
xlim	The range of the x-coordinates of the image.
ylim	The range of the y-coordinates of the image.
diggle	Logical. If TRUE, use the Jones-Diggle improved edge correction. See spatstat.explore::density.ppp() for details.

Value

Output from spatstat.explore::density.ppp.

---

Draw a contour region on some density level

Description

Draw a contour region on some density level

Usage

contour2sf(spe, contour, coi, cutoff)

Arguments

spe	A SpatialExperiment object.
contour	Name in metadata.
coi	A character vector of cell types of interest (COIs).
cutoff	A numeric scalar specifying the density cutoff.

Value

An sf object of the contour region of the specified level.

---

Test for density correlation between two cell types.

Description

Test for density correlation between two cell types.

Usage

corDensity(spe, by.roi = TRUE)

Arguments

spe	A SpatialExperiment object.
by.roi	Logical. If TRUE (default), then return the testing results at ROI level. If FALSE, then combine the testing results across all ROIs.

Value

A DataFrame containing the testing results.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts", "B cells", "T cells")

spe <- gridDensity(spe, coi = coi)

spe <- findROI(spe, coi = coi, method = "walktrap")

result <- corDensity(spe)

---

Find ROIs based on cell type-specific densities via graph-based method.

Description

Find ROIs based on cell type-specific densities via graph-based method.

Usage

findROI(
  spe,
  coi,
  probs = 0.85,
  ngrid.min = 20,
  method = "walktrap",
  diag.nodes = FALSE,
  sequential.roi.name = TRUE,
  ...
)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of cell types of interest (COIs).
probs	A numeric scalar. The threshold of proportion that used to filter grid by density. Default to 0.85.
ngrid.min	An integer. The minimum number of grids required for defining a ROI. Default to 20.
method	The community dectection method to be used, either walktrap or connected. Default to walktrap.
diag.nodes	Logical. Set this to TRUE to allow diagonal grid points to be adjacent nodes.
sequential.roi.name	Logical. Set this to FALSE if you want the original ROI name before filtering are retained.
...	Other parameters that passed to walktrap.community.

Value

A SpatialExperiment object.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts")

spe <- gridDensity(spe, coi = coi)

spe <- findROI(spe, coi = coi, method = "walktrap")

---

Get contour from density

Description

Get contour from density

Usage

getContour(spe, coi, bins = NULL, binwidth = NULL, breaks = NULL)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of cell types of interest (COIs).
bins	An integer. Number of contour levels.
binwidth	A numeric scale of the smoothing bandwidth.
breaks	A numeric scale referring to the breaks in ggplot2:::contour_breaks.

Value

A SpatialExperiment object. An sf object of the contour region of the specified level is stored in the metadata of the SpatialExperiment object.

Examples

data("xenium_bc_spe")

spe <- gridDensity(spe)

coi <- "Breast cancer"

spe <- getContour(spe, coi = coi)

---

Calculate areas between every two density levels

Description

Calculate areas between every two density levels

Usage

getContourRegions(spe, coi)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of cell types of interest (COIs).

Value

A list of sf objects, each representing the region between two contour density levels.

---

Combine grids in each ROI to a sf region

Description

Combine grids in each ROI to a sf region

Usage

grid2sf(spe)

Arguments

spe	A SpatialExperiment object.

Value

List of ROIs saved as sf objects.

---

Perform kernel density estimation on SpatialExperiment for cell types of interest

Description

Perform kernel density estimation on SpatialExperiment for cell types of interest

Usage

gridDensity(
  spe,
  coi = NULL,
  id = "cell_type",
  kernel = "gaussian",
  bandwidth = NULL,
  ngrid.x = 100,
  ngrid.y = NULL,
  grid.length.x = NULL,
  grid.length.y = NULL,
  diggle = FALSE
)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of cell types of interest (COIs). Default to all cell types.
id	A character. The name of the column of colData(spe) containing the cell type identifiers. Set to cell_type by default.
kernel	The smoothing kernel. Options are "gaussian", "epanechnikov", "quartic" or "disc".
bandwidth	The smoothing bandwidth. By default performing automatic bandwidth selection using cross-validation using function spatstat.explore::bw.diggle.
ngrid.x	Number of grids in the x-direction. Default to 100.
ngrid.y	Number of grids in the y-direction.
grid.length.x	Grid length in the x-direction.
grid.length.y	Grid length in the y-direction.
diggle	Logical. If TRUE, use the Jones-Diggle improved edge correction. See spatstat.explore::density.ppp() for details.

Value

A SpatialExperiment object. Grid density estimates for all cell type of interest are stored in spe@metadata$grid_density. Grid information is stored in spe@metadata$grid_info

Examples

data("xenium_bc_spe")

spe <- gridDensity(spe)

---

Manually merge ROIs

Description

Manually merge ROIs

Usage

mergeROI(spe, merge.list, id = "component", rename = FALSE)

Arguments

spe	A SpatialExperiment object.
merge.list	A (named) list of vectors of ROI ids to be merged. Each vector in the list should be of length greater than or equal to 2. If no name is specified, the merged ROI will be named by concatenating ROIs being merged.
id	Character. The name of the column in spe@metadata$roi that stores the ROIs to be merged. Default is "component".
rename	Logical. If TRUE, names of merge.list are ignored. ROIs will be given a new name. For the unmerged ROIs, their new names are not necessarily the same as those before merging.

Value

A SpatialExperiment object.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts")

spe <- gridDensity(spe, coi = coi)

spe <- findROI(spe, coi = coi, method = "walktrap")

spe <- mergeROI(spe, list("1-2" = 1:2))

---

Plot cell type composition in each density level of cell of interest.

Description

Plot cell type composition in each density level of cell of interest.

Usage

plotCellCompo(
  spe,
  coi,
  id = "cell_type",
  level.name = paste0(janitor::make_clean_names(coi), "_contour"),
  by.roi = FALSE
)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of cell types of interest (COIs).
id	A character. The name of the column of colData(spe) containing the cell type identifiers. Set to cell_type by default.
level.name	The column name that generated by cellAssign.
by.roi	Logical. Default is FALSE, set to TRUE to allow plotting by ROI.

Value

A ggplot object.

Examples

data("xenium_bc_spe")
spe <- gridDensity(spe, coi = c("Breast cancer", "Fibroblasts"))
spe <- findROI(spe, coi = c("Breast cancer", "Fibroblasts"))
spe <- getContour(spe, coi = "Breast cancer")
spe <- allocateCells(spe)
plotCellCompo(spe, coi = "Breast cancer")
plotCellCompo(spe, coi = "Breast cancer", by.roi = TRUE)

---

Plot contour lines.

Description

Plot contour lines.

Usage

plotContour(
  spe,
  coi,
  overlay = c("cell", "density"),
  id = "cell_type",
  sub.level = NULL,
  ...
)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of length 1 of the cell type of interest.
overlay	Character vector. Either plot overlay on density or cell. By default is cell.
id	A character. The name of the column of colData(spe) containing the cell type identifiers. Set to cell_type by default.
sub.level	Character vector. Subset on specific level.
...	Aesthetic mappings to pass to ggplot2::aes_string().

Value

A ggplot object.

Examples

data("xenium_bc_spe")

spe <- gridDensity(spe)

coi <- "Breast cancer"

spe <- getContour(spe, coi = coi)

plotContour(spe, coi = coi, size = 0.3, alpha = 0.2)

---

Visualising an sf object (for internal use only at the moment)

Description

Visualising an sf object (for internal use only at the moment)

Usage

plotContourRegion(
  spe,
  coi,
  id = "cell_type",
  overlay = c("density", "cell"),
  sub.level
)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of length 1 of the cell type of interest.
id	A character. The name of the column of colData(spe) containing the cell type identifiers. Set to cell_type by default.
overlay	Character vector. Either plot overlay on density or cells.
sub.level	Numeric vector of length 1 or 2, identifies which density level to plot. When length is 1, plot the density region above this level. When length is 2, plot the density region between the two levels.

Value

A ggplot object.

---

Plot model statistics using heatmap.

Description

Plot model statistics using heatmap.

Usage

plotCorHeatmap(
  model.result,
  stats = c("cor.coef", "t", "p.Pos", "p.Neg"),
  roi = "all",
  cell.type = "all"
)

Arguments

model.result	A data.frame object.
stats	Character value. Choose either coefficient or t. Coefficient by default.
roi	Character value. By default is all. The specific ROIs to be plotted.
cell.type	Character value. By default is all. The cell types to be plotted.

Value

A pheatmap object.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts", "B cells", "T cells")

spe <- gridDensity(spe, coi = coi)

spe <- findROI(spe, coi = coi, method = "walktrap")

model_result <- corDensity(spe)

plotCorHeatmap(model_result)

---

Plot density correlation between two cell types

Description

Plot density correlation between two cell types

Usage

plotDensCor(
  spe,
  celltype1 = NULL,
  celltype2 = NULL,
  by.roi = TRUE,
  fit = c("spline", "linear"),
  df = 3,
  ...
)

Arguments

spe	A SpatialExperiment object.
celltype1	Cell type 1 to compare.
celltype2	Cell type 2 to compare.
by.roi	Logical. Plot facet by ROIs or not.
fit	Character. Options are "spline" and "linear".
df	Integer. Degrees of freedom of the spline fit. Default to 3 (i.e., a cubic spline fit).
...	aesthetic mappings to pass to ggplot2::aes().

Value

A ggplot object.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts")

spe <- gridDensity(spe, coi = coi)

spe <- findROI(spe, coi = coi, method = "walktrap")

plotDensCor(spe, celltype1 = "Breast cancer", celltype2 = "Fibroblasts")

---

Plot grid-based density.

Description

Plot grid-based density.

Usage

plotDensity(spe, coi, probs = 0.8)

Arguments

spe	A SpatialExperiment object.
coi	A character vector of cell types of interest (COIs).
probs	Numeric value between 0 and 1, used for filtering uninformative grid, default is 0.8.

Value

A ggplot object.

Examples

data("xenium_bc_spe")

spe <- gridDensity(spe)

plotDensity(spe, coi = "Breast cancer")

plotDensity(spe, coi = "Fibroblasts")

---

Plot ROIs on spatial.

Description

Plot ROIs on spatial.

Usage

plotROI(spe, id = "cell_type", show.legend = FALSE, ...)

Arguments

spe	A SpatialExperiment object.
id	Character. The name of the column of colData(spe) containing the cell type identifiers. Set to cell_type by default.
show.legend	Logical. Show legend or not.
...	Aesthetic mappings pass for point.

Value

A ggplot object.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts")

spe <- gridDensity(spe, coi = coi)

spe <- findROI(spe, coi = coi, method = "walktrap", steps = 5)

plotROI(spe, size = 0.3, alpha = 0.2)

---

Plot cells based on spatial coordinates.

Description

Plot cells based on spatial coordinates.

Usage

plotSpatial(spe, reverseY = FALSE, n = 30, ...)

Arguments

spe	A SpatialExperiment object.
reverseY	Reverse y coordinates.
n	Integer value. The number of distinct color to be generated, default is 30.
...	Aesthetic mappings to pass to ggplot2::aes_string().

Value

A ggplot object.

Examples

data("xenium_bc_spe")

plotSpatial(spe, shape = ".", color = cell_type, size = 0.3, alpha = 0.2)

---

Merge sel_region from the selectRegion function to SpatialExperiment.

Description

Merge sel_region from the selectRegion function to SpatialExperiment.

Usage

postSelRegion(spe, sel_region)

Arguments

spe	A SpatialExperiment object.
sel_region	A dataframe object. Can be generated from function selectRegion.

Value

A SpatialExperiment object.

Examples

data("xenium_bc_spe")

coi <- c("Breast cancer", "Fibroblasts", "B cells", "T cells")

spe <- gridDensity(spe, coi = coi)

sel_region <- data.frame(
    "node" = seq(10),
    "node_x" = seq(10),
    "node_y" = seq(10)
)

spe1 <- postSelRegion(spe, sel_region)

---

Select region of interest from plot

Description

Select region of interest from plot

Usage

selectRegion(data, x.col = "x", y.col = "y")

Arguments

data	A data.frame object.
x.col	Column name of the x coordinates.
y.col	Column name of the y coordinates.

Value

A data.frame object in the global environment.

Examples

data("xenium_bc_spe")

spe_b <- spe[, SummarizedExperiment::colData(spe)$cell_type == "B cells"]

dat <- as.data.frame(SpatialExperiment::spatialCoords(spe_b))

# selectRegion(dat, x.col = "x_centroid", y.col = "y_centroid")

---

Given a 'SpatialExperiment' data object, create pseudo-bulk samples using the colData information and return a DGEList object

Description

Given a 'SpatialExperiment' data object, create pseudo-bulk samples using the colData information and return a DGEList object

Usage

spe2PB(
  spe,
  by.group = TRUE,
  group.id = "cell_type",
  by.roi = TRUE,
  roi.only = TRUE,
  contour = NULL
)

Arguments

spe	A SpatialExperiment object.
by.group	Logical. Whether to perform pseudo-bulking by group. TRUE by default.
group.id	Character. The column name of the colData(spe) that contains the group information. Default to 'cell_type'.
by.roi	Logical. Whether to perform pseudo-bulking by ROI. TRUE by default.
roi.only	Logical. Whether to filter out pseudo-bulk samples formed by cells not in any ROIs. TRUE by default.
contour	Character. The name of the group or cell type on which the contour level is computed. If NULL, then no pseudo-bulking will be performed based on contour level. Default to NULL.

Value

An edgeR::DGEList object where each library (column) is a pseudo-bulk sample.

Examples

data("xenium_bc_spe")

spe <- gridDensity(spe)

coi <- "Breast cancer"

spe <- findROI(spe, coi = coi)

spe <- allocateCells(spe)

y <- spe2PB(spe)

---