add spatial clustering process in simulated dataset

src/metrics/downstream/script.R

@@ -41,11 +41,11 @@ crosscor_mantel <- generate_mantel(real_moransI, sim_moransI)
 
 cat("spatial clustering evaluation\n")
 # TODO
-# sim_sce <- scater::logNormCounts(SingleCellExperiment::SingleCellExperiment(
-#   list(counts = Matrix::t(input_simulated_sp$layers[["counts"]])),
-#   colData = input_simulated_sp$obs,
-#   metadata = input_simulated_sp$obsm
-# ))
+sim_sce <- scater::logNormCounts(SingleCellExperiment::SingleCellExperiment(
+  list(counts = Matrix::t(input_simulated_sp$layers[["counts"]])),
+  colData = input_simulated_sp$obs,
+  metadata = input_simulated_sp$obsm
+))
 
 # generate the simulated clustering result first by BayersSpace
 # sim_sce <- BayesSpace::spatialPreprocess(sim_sce, platform=par$plat, 

src/metrics/ks_statistic_spatial/script.py

@@ -102,8 +102,7 @@ def get_trans(adata=None, ct=None):
 
 # transition matrix
 real = np.array(input_spatial_dataset.obs['spatial_cluster'].values.tolist())
-# sim = np.array(input_simulated_dataset.obs['spatial_cluster'].values.tolist())
-sim <- generate_sim_spatialCluster(input_spatial_dataset, input_simulated_dataset)
+sim = np.array(input_simulated_dataset.obs['spatial_cluster'].values.tolist())
 
 
 transition_matrix_real = get_trans(adata=input_spatial_dataset, ct=real)

src/process_datasets/generate_sim_spatialcluster/config.vsh.yaml

@@ -0,0 +1,55 @@
+name: generate_sim_sparialcluster
+namespace: process_datasets
+description: generate spatial cluster of simulated dataset before run metrics folder
+argument_groups:
+  - name: Inputs
+    arguments:
+      - type: file
+        name: --input_sp
+        description: Raw spatial dataset
+        example: resources_test/datasets/MOBNEW/dataset_sp.h5ad
+        required: true
+      - type: file
+        name: --input_sp_sim
+        description: similated spatial dataset
+        example: resources_test/datasets/MOBNEW/simulated_dataset.h5ad
+        required: true
+
+  - name: Outputs
+    arguments:
+      - type: file
+        name: --output_sp
+        description: Processed simulated spatial dataset
+        example: simulated_dataset.h5ad
+        direction: output
+        required: true
+
+resources:
+  - type: r_script
+    path: script.R
+  - path: /src/helpers/utils.R
+
+test_resources:
+  - path: /resources_test/datasets/MOBNEW
+    dest: resources_test/datasets/MOBNEW
+  - type: python_script
+    path: /common/component_tests/run_and_check_output.py
+
+engines:
+  - type: docker
+    image: openproblems/base_r:1.0.0
+    setup:
+      - type: r
+        bioc: 
+          - scater
+          - BayesSpace
+          - SingleCellExperiment
+        cran:
+          - aricode
+          - anndata
+          - reshape2
+
+
+runners:
+  - type: executable
+  - type: nextflow

src/process_datasets/generate_sim_spatialcluster/script.R

@@ -0,0 +1,37 @@
+
+
+## VIASH START
+par <- list(
+  # inputs
+  input_spatial_dataset = "resources_test/datasets/MOBNEW/dataset_sp.h5ad",
+  input_simulated_dataset = "resources_test/datasets/MOBNEW/simulated_dataset.h5ad",
+  # outputs
+  output_sp = "resources_test/datasets/MOBNEW/simulated_dataset.h5ad"
+)
+meta <- list(
+  resources_dir = "target/executable/process_datasets/generate_sim_sparialCluster"
+)
+## VIASH END
+source(file.path(meta$resources_dir, "utils.R"))
+
+cat("Read input files\n")
+input_real_sp <- anndata::read_h5ad(par$input_spatial_dataset)
+input_simulated_sp <- anndata::read_h5ad(par$input_simulated_dataset)
+
+sim_sce <- scater::logNormCounts(SingleCellExperiment::SingleCellExperiment(
+  list(counts = Matrix::t(input_simulated_sp$layers[["counts"]])),
+  colData = input_simulated_sp$obs,
+  metadata = input_simulated_sp$obsm
+))
+cat("add spatial cluster in simulated dataset:\n")
+sim_cluster <- generate_sim_spatialCluster(input_real_sp, input_simulated_sp)
+
+# need reclassify again
+real_cluster <- input_real_sp$obs[,c("spatial_cluster")]
+location <- colnames(counts(sim_sce))
+sim_new_cluster <- reclassify_simsce(location, real_cluster, sim_cluster)
+
+input_simulated_sp$obs$spatial_cluster <- sim_new_cluster
+
+cat("Writing output to file\n")
+input_simulated_sp$write_h5ad(par$output_sp, compression = "gzip")