Run CytoSPACE spatial assignment

Usage

RunCytoSPACE(
  srt,
  reference,
  reference_label,
  assay = NULL,
  reference_assay = NULL,
  layer = "counts",
  reference_layer = "counts",
  features = NULL,
  cell_fractions = NULL,
  n_cells_per_spot = NULL,
  mean_cell_numbers = 5,
  scRNA_max_transcripts_per_cell = 1500,
  sampling_method = "duplicates",
  seed = 1,
  prefix = "CytoSPACE",
  store_results = TRUE,
  verbose = TRUE,
  image = NULL,
  coord.cols = c("col", "row"),
  coordinate_space = c("raw", "legacy_display"),
  backend = c("cpp", "r"),
  max_dense_gib = 8
)

Arguments

srt: A Seurat object.
reference: Reference Seurat object containing annotated single cells.
reference_label: Metadata column in reference with cell type labels.
assay: Assay to use. NULL uses the default assay.
reference_assay: Assay used in reference.
layer, reference_layer: Assay layers used for spatial and reference expression.
features: Features used for assignment. If NULL, shared features are used.
cell_fractions: Optional cell-type fractions. Provide a named numeric vector, one-row matrix/data.frame, or a spot-by-cell-type matrix/data.frame. Spot-level rows are aggregated to the global composition used by the default CytoSPACE assignment workflow.
n_cells_per_spot: Optional number of cells assigned to each spatial spot. If NULL, counts are estimated from spatial RNA reads with mean_cell_numbers.
mean_cell_numbers: Mean number of cells per spot. Default 5, matching the CytoSPACE Visium default.
scRNA_max_transcripts_per_cell: Maximum reference transcripts per cell before assignment. Default 1500, matching CytoSPACE.
sampling_method: Sampling method. Only "duplicates" is supported in the package runtime.
seed: Random seed used for deterministic reference downsampling and duplicate sampling.
prefix: Prefix for metadata columns.
store_results: Whether to store detailed assignment results in srt@tools.
verbose: Whether to print the message. Default is TRUE.
image: Optional Seurat image used for spatial coordinates.
coord.cols: Metadata coordinate columns used when no image is selected.
coordinate_space: Coordinate space used for assignment locations. The default is raw acquisition coordinates. Use "legacy_display" explicitly to reproduce the display-scaled locations used before scop 0.9.0.
backend: Numerical backend used to estimate cell-type fractions when cell_fractions is not supplied. "cpp" fuses normalization, reference centroid construction, correlation, and weighted aggregation; "r" keeps the reference implementation. Spot assignment uses C++ in both cases.
max_dense_gib: Maximum estimated GiB allowed for dense expression working matrices.

Value

A Seurat object with CytoSPACE metadata columns and detailed results stored in srt@tools[["CytoSPACE"]].

Examples

data(visium_human_pancreas_sub)
data(panc8_sub)
spatial <- visium_human_pancreas_sub[, seq_len(120)]
#> Warning: Not validating Centroids objects
#> Warning: Not validating Centroids objects
#> Warning: Not validating FOV objects
#> Warning: Not validating FOV objects
#> Warning: Not validating FOV objects
#> Warning: Not validating FOV objects
#> Warning: Not validating FOV objects
#> Warning: Not validating FOV objects
#> Warning: Not validating Seurat objects
reference <- panc8_sub[, panc8_sub$celltype %in% c("ductal", "alpha", "beta")]
reference <- Seurat::FindVariableFeatures(reference, nfeatures = 300, verbose = FALSE)
features_use <- intersect(
  SeuratObject::VariableFeatures(reference),
  rownames(spatial)
)
spatial <- RunCytoSPACE(
  spatial,
  reference = reference,
  reference_label = "celltype",
  features = features_use,
  mean_cell_numbers = 1,
  verbose = FALSE
)

SpatialSpotPlot(
  visium_human_pancreas_sub,
  group.by = "coda_label",
  theme_use = "theme_scop"
)


SpatialSpotPlot(
  spatial,
  group.by = "CytoSPACE_dominant_type",
  theme_use = "theme_scop"
)