Run mistyR multiview spatial modeling

Build a small mistyR view composition from a spatial Seurat object, train MISTy models, collect results, and store a standardized result bundle in srt@tools. The intraview is always created from the selected assay layer; optional juxtaview and paraview components describe local and broader spatial context. mistyR is an optional Bioconductor dependency installable with BiocManager::install("mistyR").

Usage

RunMistyR(
  srt,
  assay = NULL,
  layer = "data",
  features = NULL,
  image = NULL,
  coord.cols = c("col", "row"),
  views = "para",
  para_l = 10,
  para_zoi = 0,
  para_family = c("gaussian", "exponential", "linear", "constant"),
  para_approx = 1,
  para_nn = NULL,
  juxta_neighbor_thr = 15,
  view_cached = FALSE,
  results_folder = NULL,
  seed = 42,
  target_subset = NULL,
  bypass_intra = FALSE,
  cv_folds = 10,
  model_cached = FALSE,
  append = FALSE,
  tool_name = "MistyR",
  store_results = TRUE,
  store_views = FALSE,
  verbose = TRUE,
  coordinate_space = c("raw", "legacy_display"),
  ...
)

Arguments

srt: A Seurat object.
assay: Assay used for expression. If NULL, the default assay is used.
layer: Assay layer used for expression values.
features: Features used by MISTy. If NULL, variable features are used when available; otherwise all assay features are used.
image: Name of the Seurat spatial image. Required when multiple images are present; a single image is selected automatically when NULL.
coord.cols: Metadata coordinate columns used when no Seurat image coordinates are available.
views: Spatial views to add besides the required intraview. One or both of "para" and "juxta".
para_l, para_zoi, para_family, para_approx, para_nn: Parameters passed to mistyR::add_paraview(). para_l and para_zoi use the selected coordinate units; para_nn is a unitless neighbor count.
juxta_neighbor_thr: Neighbor threshold passed to mistyR::add_juxtaview(), expressed in the selected coordinate units.
view_cached: Whether generated mistyR views should use cache.
results_folder: Folder passed to mistyR::run_misty(). If NULL, a temporary folder is used.
seed, target_subset, bypass_intra, cv_folds, model_cached, append: Parameters passed to mistyR::run_misty().
tool_name: Name used to store results in srt@tools.
store_results: Whether to store results in srt@tools.
store_views: Whether to store the mistyR view composition in srt@tools. This can be large.
verbose: Whether to print the message. Default is TRUE.
coordinate_space: Coordinate system used to build MISTy views. The default is raw acquisition coordinates; "legacy_display" remains an explicit compatibility option.
...: Additional named arguments passed to mistyR::run_misty().

Value

A Seurat object with results stored in srt@tools[[tool_name]] when store_results = TRUE.

Examples

data(visium_human_pancreas_sub)
spatial <- visium_human_pancreas_sub
spatial <- Seurat::NormalizeData(spatial, assay = "Spatial", verbose = FALSE)

spatial <- RunMistyR(
  spatial,
  assay = "Spatial",
  layer = "data",
  features = rownames(spatial)[1:10],
  coord.cols = c("x", "y"),
  views = "para",
  para_l = 5,
  cv_folds = 3,
  verbose = FALSE
)
#> 
#> Training models
#> Registered S3 method overwritten by 'pROC':
#>   method   from            
#>   plot.roc spatstat.explore
#> 
#> Collecting improvements
#> 
#> Collecting contributions
#> 
#> Collecting importances
#> 
#> Aggregating
spatial@tools$MistyR$summary
#> $views
#> [1] "intraview"      "misty.uniqueid" "paraview.5"    
#> 
#> $n_targets
#> [1] 10
#> 
#> $n_improvement_records
#> [1] 80
#> 
#> $n_contribution_records
#> [1] 60
#> 
#> $importance_views
#> [1] "view"       "Predictor"  "Target"     "Importance" "nsamples"  
#>