Merge pull request #41 from lorenzobonaguro/dev

GitHub update to Version 0.3.1

Author: lorenzobonaguro

DESCRIPTION

@@ -62,7 +62,8 @@ Imports:
     rstatix,
     tidyr,
     CytoNorm,
-    corrplot
+    corrplot,
+    S4Vectors
 Remotes:
     JinmiaoChenLab/Rphenograph,
     stuchly/Rphenoannoy,

NAMESPACE

@@ -6,11 +6,13 @@ export(change_param_name)
 export(checkInput)
 export(check_IDs)
 export(clr)
+export(condor_info)
 export(condor_session_info)
 export(confusionMatrix)
 export(corr_plot_comparison)
 export(create_metaclustering_script)
 export(df_frequency)
+export(export_sce)
 export(filter_fcd)
 export(frequency_anova_test)
 export(frequency_friedman_test)
@@ -25,6 +27,7 @@ export(marker_wilcox_test)
 export(measured_markers)
 export(merge_condor)
 export(metaclustering)
+export(order_param)
 export(plot_confusion_HM)
 export(plot_counts_barplot)
 export(plot_dim_density)
@@ -109,6 +112,10 @@ import(umap)
 import(uwot)
 importFrom(Biobase,AnnotatedDataFrame)
 importFrom(Matrix,sparseMatrix)
+importFrom(S4Vectors,metadata)
+importFrom(SingleCellExperiment,SingleCellExperiment)
+importFrom(SingleCellExperiment,reducedDim)
+importFrom(SummarizedExperiment,colData)
 importFrom(grDevices,recordPlot)
 importFrom(igraph,membership)
 importFrom(utils,packageDescription)

R/astir.R

@@ -10,6 +10,7 @@
 #' @param max_epochs Maximum number of epochs, for details see `Astir` documentation.
 #' @param learning_rate Learning Rate.
 #' @param initial_epochs initial epochs, for details see `Astir` documentation.
+#' @param prefix Optional prefix for the slot name of the output.
 #' @details Predict cell types using 'Astir'. This package requires the python library `astir` and `reticulate`. This function is still experimental and was not extensively tested in cyCONDOR. In one of the next release we will document it's usage in more details on a dedicated vignette.
 #' For the moment if you want to know more feel free to reach out on our `Slack` channel. For more details on the different parameters refer to `Astir` documentation.
 #' @import reticulate
@@ -23,7 +24,7 @@ runAstir_celltype <- function(fcd,
                               max_epochs,
                               learning_rate,
                               initial_epochs,
-                              prefix = "") {
+                              prefix = NULL) {
 
   # Save the expression matrix as csv
   write.csv(x = fcd$expr[[data_slot]], file = paste0(analysis_path, "expr.csv"))
@@ -68,7 +69,7 @@ runAstir_celltype <- function(fcd,
                                         "Learning_Rate_", learning_rate,
                                         "_Initial_Epochs_", initial_epochs))
 
-  fcd[["astir"]][[paste(prefix, "Astir_cell_type", data_slot, sep = "_")]] <- df
+  fcd[["astir"]][[paste("Astir_cell_type", sub("^_", "", paste(prefix, data_slot, sep = "_")), sep = "_")]] <- df
 
   return(fcd)
 
@@ -86,6 +87,7 @@ runAstir_celltype <- function(fcd,
 #' @param max_epochs Maximum number of epochs.
 #' @param learning_rate Learning Rate.
 #' @param initial_epochs Initial Epochs.
+#' @param prefix Optional prefix for the slot name of the output.
 #' @import reticulate
 #' @importFrom utils write.csv
 #' @return runAstir_cellstates
@@ -98,7 +100,7 @@ runAstir_cellstates <- function(fcd,
                                 max_epochs,
                                 learning_rate,
                                 initial_epochs,
-                                prefix = "") {
+                                prefix = NULL) {
 
   # Save the expression matrix as csv
   write.csv(x = fcd$expr[[data_slot]], file = paste0(analysis_path, "expr.csv"))
@@ -138,7 +140,7 @@ runAstir_cellstates <- function(fcd,
                                         "Learning_Rate_", learning_rate,
                                         "_Initial_Epochs_", initial_epochs))
 
-  fcd[["astir"]][[paste(prefix, "Astir_cell_state", data_slot, sep = "_")]] <- df
+  fcd[["astir"]][[paste("Astir_cell_state", sub("^_", "", paste(prefix, data_slot, sep = "_")), sep = "_")]] <- df
 
   return(fcd)
 

R/data_load_and_transform.R

@@ -656,3 +656,80 @@ write_fcs <- function(fcd = condor,
     message("File was saved to ", dir, filename, ".fcs")
   }
 }
+
+#' Save the results from cyCONDOR as a Single Cell Experiment Class as RDS file
+#'
+#' @title Export Flow Cytometry Dataset as SCE.
+#' @description This function creates an object of the class Single Cell Experiment (SCE) based on the data of the flow cytometry dataset (fcd).
+#' @param fcd flow cytometry data set
+#' @param expr_slot expr_slot from which to take marker expression values, default is "orig".
+#' @param reduction_method string specifying which dimensionality reduction method to use.
+#' @param reduction_slot string specifying reduction name in reduction_method to use for visualization, e.g. "pca_orig".
+#' @param cluster_slot string specifying which clustering slot to use to find variable specified in cluster_var.
+#' @param cluster_var string specifying variable name in cluster_slot that identifies cell population labels to be used (e.g. clusters, metaclusters or predicted labels). Must be numeric in this function - factors are converted automatically.
+#' @param save boolean indicating if the sce should be saved as RDS file.
+#' @param dir string specifying the directory where the RDS file is saved. Current working directory by default.
+#' @param filename string specifying the filename for the RDS file.
+#' @importFrom SingleCellExperiment SingleCellExperiment reducedDim
+#' @importFrom SummarizedExperiment colData
+#' @importFrom S4Vectors metadata
+#' @returns The SCE  plus a short message if the RDS file was saved successfully.
+#'
+#' @export
+export_sce <- function(fcd = condor,
+                       expr_slot = "orig",
+                       reduction_method = NULL,
+                       reduction_slot = NULL,
+                       cluster_slot = NULL,
+                       cluster_var = NULL,
+                       save = TRUE,
+                       dir = paste0(getwd(), "/"),
+                       filename = "") {
+
+  message("This function is still experimental, if you run into any issue please let us know!")
+
+  # Extracting the expression data
+  exp <- as.matrix(t(fcd[["expr"]][[expr_slot]]))
+
+  # Extracting the annotation data
+  anno <- fcd[["anno"]][["cell_anno"]]
+  names(anno)[names(anno) == "sample_ID"] <- "sample_id" # required for CATALYST - .check_sce() checks if 'sample_id' (case sensitive!) is present in the SCE.
+  names(anno)[names(anno) == "group"] <- "condition" # also required naming for CATALYST
+
+  # Check if the colnames of exp and the rownames of anno match
+  if(!identical(colnames(exp), rownames(anno))){
+    stop("Column names of the expression matrix and rownames of the annotation matrix do not match.")
+  }
+
+  # Create the SCE
+  sce <- SingleCellExperiment::SingleCellExperiment(assays = list(assays = exp),
+                                                    colData = anno)
+
+  # Extract and add dimensional reduction data
+  if(!is.null(reduction_method)){
+    red <- fcd[[reduction_method]][[reduction_slot]]
+    SingleCellExperiment::reducedDim(sce, reduction_method) <- red
+  }
+
+  # Extract and add clustering information
+  if(!is.null(cluster_slot)){
+    clust <- fcd[["clustering"]][[cluster_slot]][[cluster_var]]
+    SummarizedExperiment::colData(sce)[[cluster_slot]] <- clust
+    SummarizedExperiment::colData(sce)[["cluster_id"]] <- clust # also required for downstream application in CATALYST
+  }
+
+  # Extract and save experiment_info as metadata
+  ei <- unique(as.data.frame(colData(sce)[, c("sample_id", "condition")]))
+  rownames(ei) <- ei$sample_id  # Set rownames = sample IDs
+  S4Vectors::metadata(sce)$experiment_info <- ei
+
+  # Save the SCE
+  ## as .rds for standard usage in R
+  if(save == TRUE) {
+    saveRDS(sce,
+            file = paste0(dir, filename, ".rds"))
+    message("SCE has been saved to ", dir, filename, ".rds")
+  }
+
+  return(sce)
+}

R/data_visualization.R

@@ -225,6 +225,8 @@ confusionMatrix <- function (i = NULL, j = NULL)
 #' @param cluster_slot string specifying which clustering slot to use to find variable specified in cluster_var.
 #' @param add_pseudotime Logical, if plot should be colored by pseudotime.
 #' @param pseudotime_slot string specifying pseudotime name to use for visualization.
+#' @param add_astir Logical, if plot should be colored by Astir prediction.
+#' @param astir_slot string specifying astir description name to use for visualization.
 #' @param param parameter to visualize in the plot, this can be either a continuous variable or a categorical one, the function will react differently accordingly.
 #' @param order logical if you want to order the dots in the plot, by expression for example. This can help to find small populations of positive cells. If set to FALSE, the plotting order of the cells is randomized.
 #' @param title title of the plot.

_pkgdown.yml

@@ -104,6 +104,9 @@ reference:
   - condor_session_info
   - corr_plot_comparison
   - write_fcs
+  - condor_info
+  - export_sce
+  - order_param
 
 articles:
 - title: Detailed cyCONDOR workflows
@@ -129,5 +132,5 @@ articles:
 - title: Other workflows
   desc: >
     Additional analytical workflows
-  contents: 
+  contents:
   - Workflow_merge_panels