Add mesenchymal dataset

code/download-raw-data.R

@@ -147,3 +147,15 @@ DimPlot(cbmc, label = TRUE)
 citeseq <- as.SingleCellExperiment(cbmc)
 
 saveRDS(citeseq, here::here("data", "raw_data", "citeseq.rds"))
+
+# Endothelial
+
+endothelial <- readRDS(here::here("data", "downloaded_data",
+                                  "20201104_endothelial_annotated.rds" ))
+saveRDS(endothelial, here::here("data", "raw_data", "endothelial.rds"))
+
+# Mesenchymal
+
+mesenchymal <- readRDS(here::here("data", "downloaded_data",
+                                  "20201215_Mesenchymal_updated_annotation.rds"))
+saveRDS(mesenchymal, here::here("data", "raw_data", "mesenchymal.rds"))

code/prep-mesenchymal.R

+# This scripts prepares and saves the `mesenchymal` data.
+
+args <- R.utils::commandArgs(
+  trailingOnly = TRUE,
+  asValues = TRUE
+)
+
+suppressMessages({
+  library(scater)
+  library(scran)
+  library(Seurat)
+})
+
+set.seed(31012022)
+
+mesenchymal <- readRDS(args$raw_data)
+
+# Remove unneeded assays which prevent conversion to a SCE object.
+mesenchymal@assays$HTO <- NULL
+mesenchymal@assays$SCT <- NULL
+
+mesenchymal <- as.SingleCellExperiment(mesenchymal)
+
+rowData(mesenchymal) <- DataFrame(gene = rownames(mesenchymal))
+
+colData(mesenchymal)$predicted.id <- factor(colData(mesenchymal)$predicted.id)
+colLabels(mesenchymal) <- colData(mesenchymal)$predicted.id
+
+clusters <- quickCluster(mesenchymal)
+mesenchymal <- computeSumFactors(mesenchymal, clusters = clusters)
+mesenchymal <- logNormCounts(mesenchymal)
+
+dec_mesenchymal <- modelGeneVarByPoisson(mesenchymal)
+top_mesenchymal <- getTopHVGs(dec_mesenchymal, n = 2000)
+mesenchymal <- mesenchymal[top_mesenchymal, ]
+
+saveRDS(mesenchymal, args$data)

config.yaml

@@ -11,6 +11,7 @@ all_data_ids:
   - "zhao"
   - "ss3_pbmc"
   - "citeseq"
+  - "mesenchymal"
 general_sim_data_ids: ["zeisel", "pbmc3k", "lawlor", "paul"]
 time_sim_data_ids: ["pbmc3k"]