Run SpatialEcoTyper spatial ecotype analysis

Run SpatialEcoTyper workflows through the optional SpatialEcoTyper package and write spatial ecotype labels or abundances back to a Seurat object when possible.

Usage

RunSpatialEcoTyper(
  srt,
  mode = c("single", "multi", "recover", "deconvolute"),
  assay = NULL,
  layer = "data",
  celltype.by = NULL,
  sample.by = NULL,
  x.by = "X",
  y.by = "Y",
  dat = NULL,
  celltypes = NULL,
  features = NULL,
  outprefix = NULL,
  outdir = NULL,
  radius = 50,
  resolution = 0.5,
  nfeatures = 300,
  min.cts.per.region = 2,
  npcs = 20,
  min.cells = 5,
  min.features = 10,
  iterations = 10,
  minibatch = 5000,
  ncores = 4,
  grid.size = round(radius * 1.4),
  filter.region.by.celltypes = NULL,
  k = 20,
  k.sn = 50,
  dropcell = FALSE,
  normalization.method = "None",
  nmf_ranks = 10,
  nrun.per.rank = 30,
  min.coph = 0.95,
  Region = NULL,
  downsample.by.region = TRUE,
  subresolution = 30,
  seed = 1,
  scale = TRUE,
  Ws = NULL,
  ncell.per.run = 500,
  min.score = 0.6,
  W = NULL,
  nsample.per.run = 500,
  sum2one = TRUE,
  prefix = "SpatialEcoTyper",
  tool_name = "SpatialEcoTyper",
  store_results = TRUE,
  allow_partial = FALSE,
  verbose = TRUE,
  ...
)

Arguments

srt

A Seurat object. For mode = "deconvolute", a numeric expression matrix can also be supplied.

mode

SpatialEcoTyper workflow. "single" runs single-sample de novo discovery, "multi" runs conserved ecotype discovery across samples, "recover" recovers pretrained SE labels, and "deconvolute" infers SE abundances from bulk or spot-level expression.

assay

Assay used for expression extraction. If NULL, the default assay is used.

layer

Assay layer used for expression extraction.

celltype.by

Metadata column containing cell type annotations. Required for "single", "multi", and "recover" unless celltypes is supplied for "recover".

sample.by

Metadata column identifying samples for mode = "multi".

x.by, y.by

Metadata columns containing single-cell spatial coordinates.

dat

Optional expression matrix used by "recover" or "deconvolute". If NULL, expression is extracted from srt.

celltypes

Optional named vector of cell types passed to SpatialEcoTyper::RecoverSE().

features

Optional feature vector used to subset the expression matrix.

outprefix

Output prefix passed to SpatialEcoTyper. Use NULL to avoid writing single-sample result files to the working directory.

outdir

Output directory passed to multi-sample SpatialEcoTyper. NULL creates a temporary directory.

radius

Spatial neighborhood radius, in the same units as x.by and y.by.

resolution

Louvain clustering resolution used by SpatialEcoTyper.

nfeatures

Number of variable features used by SpatialEcoTyper.

min.cts.per.region

Minimum number of cell types required in a spatial neighborhood.

npcs

Number of principal components used for similarity networks.

min.cells

Minimum number of cells or spatial meta-cells expressing a feature.

min.features

Minimum number of features detected in a cell or spatial meta-cell.

iterations

Number of similarity network fusion iterations.

minibatch

Number of columns processed per mini-batch in SNF.

ncores

Number of CPU cores used by SpatialEcoTyper.

grid.size

Spatial grid size used to discretize coordinates.

filter.region.by.celltypes

Optional cell types used to restrict spatial neighborhoods.

k

Number of spatial nearest neighbors used to construct spatial meta-cells.

k.sn

Number of nearest neighbors used to construct similarity networks.

dropcell

Whether cells without spatial ecotype assignments are removed from the returned SpatialEcoTyper metadata.

normalization.method, nmf_ranks, nrun.per.rank, min.coph, Region, downsample.by.region, subresolution, seed

Parameters passed to SpatialEcoTyper::MultiSpatialEcoTyper().

scale

Whether to scale expression for "recover" and "deconvolute".

Ws

Pretrained basis matrices passed to SpatialEcoTyper::RecoverSE().

ncell.per.run

Number of cells processed per run by SpatialEcoTyper::RecoverSE().

min.score

Minimum prediction score passed to SpatialEcoTyper::RecoverSE().

W

Pretrained basis matrix passed to SpatialEcoTyper::DeconvoluteSE().

nsample.per.run

Number of samples processed per run by SpatialEcoTyper::DeconvoluteSE().

sum2one

Whether inferred SE abundances are normalized to sum to one.

prefix

Prefix used for output metadata columns.

tool_name

Name used to store detailed results in srt@tools.

store_results

Whether to store raw results in srt@tools.

allow_partial

Whether to allow missing SE labels for cells absent from returned SpatialEcoTyper metadata. Default is FALSE to avoid silent partial annotations.

verbose

Whether to print the message. Default is TRUE.

...

Additional arguments passed to the selected SpatialEcoTyper function.

Value

A Seurat object with SpatialEcoTyper results in metadata and raw results stored in srt@tools[[tool_name]] when store_results = TRUE. For matrix input with mode = "deconvolute", the abundance matrix is returned.

Examples

data(visium_human_pancreas_sub)
srt <- visium_human_pancreas_sub
srt$CellType <- srt$coda_label
srt$SpatialEcoTyper_SE <- ifelse(srt$x > stats::median(srt$x), "SE1", "SE2")
srt$sample <- ifelse(srt$y > stats::median(srt$y), "slice_a", "slice_b")

SpatialEcoTyperSpatialPlot(
  srt,
  overlay_image = FALSE,
  coord.cols = c("x", "y")
)

SpatialEcoTyperCompositionPlot(
  srt,
  group.by = "CellType",
  sample.by = "sample",
  position = "fill"
)


if (
  isTRUE(check_r("digitalcytometry/SpatialEcoTyper", verbose = FALSE))
) {
srt <- RunSpatialEcoTyper(
  srt,
  celltype.by = "CellType",
  x.by = "x",
  y.by = "y",
  nfeatures = 100,
  ncores = 1,
  verbose = FALSE
)

srt <- RunSpatialEcoTyper(
  srt,
  mode = "multi",
  celltype.by = "CellType",
  sample.by = "sample",
  x.by = "x",
  y.by = "y",
  nfeatures = 100,
  ncores = 1,
  verbose = FALSE
)
}
#> Error in check_r("digitalcytometry/SpatialEcoTyper", verbose = FALSE): could not find function "check_r"