Merge pull request #29 from lorenzobonaguro/dev

Version 0.3.0

Author: lorenzobonaguro

.Rhistory

@@ -1,5 +1,4 @@
-condor.Rproj
+cyCONDOR.Rproj
 .Rproj.user/
-inst/doc
 .test_files
 .Rhistory

.gitignore

@@ -1,19 +1,21 @@
 Package: cyCONDOR
 Type: Package
 Title: Flow Cytometry data analysis toolbox
-Version: 0.2.1
+Version: 0.3.0
 Authors@R: c(
   person(given = "Lorenzo", family = "Bonaguro", email = "lorenzo.bonaguro@dzne.de", role = c("aut", "cre")),
   person(given = "Charlotte", family = "Kroeger", email = "charlotte.kroeger@dzne.de", role = "aut"),
   person(given = "Sophie", family = "Mueller", email = "sophie.mueller@dzne.de", role = "aut"),
-  person(given = "Jacqueline", family = "Leidner", email = "jacqueline.leidner@dzne.de", role = "aut")
+  person(given = "Jacqueline", family = "Leidner", email = "jacqueline.leidner@dzne.de", role = "aut"),
+  person(given = "Aleksej", family = "Frolov", email = "aleksej.frolov@dzne.de", role = "aut"),
+  person(given = "Francesco", family = "Elli", email = "francesco.elli@dzne.de", role = "aut"),
+  person(given = "Nico", family = "Henschel", email = "nico.henschel@dzne.de", role = "aut")
   )
-Description: Flow cytometry analysis workflow. The aim of this project if to 
+Description: Flow cytometry analysis workflow. The aim of this project is to 
     provide an intuitive workflow for the analysis of high-dimensionality cytometry data in R.
-URL: https://github.com/lorenzobonaguro/condor
+URL: https://lorenzobonaguro.github.io/cyCONDOR
 License: GPL-3 + file LICENSE
 Encoding: UTF-8
-LazyData: FALSE
 biocViews:
 Imports: 
     ggplot2, 
@@ -58,9 +60,17 @@ Imports:
     ggridges,
     rstatix,
     tidyr,
-    CytoNorm
-RoxygenNote: 7.2.3
+    CytoNorm,
+    corrplot
+Remotes:
+    JinmiaoChenLab/Rphenograph,
+    stuchly/Rphenoannoy,
+    saeyslab/CytoNorm@362ac08
+RoxygenNote: 7.3.2
 VignetteBuilder: knitr
 Suggests: 
     rmarkdown,
-    knitr
+    knitr,
+    testthat (>= 3.0.0),
+    here
+Config/testthat/edition: 3

DESCRIPTION

@@ -1,11 +1,14 @@
 # Generated by roxygen2: do not edit by hand
 
 export(PC_loadings)
+export(add_diffcyt_statistics)
 export(change_param_name)
 export(checkInput)
 export(check_IDs)
 export(clr)
+export(condor_session_info)
 export(confusionMatrix)
+export(corr_plot_comparison)
 export(create_metaclustering_script)
 export(df_frequency)
 export(filter_fcd)
@@ -53,13 +56,15 @@ export(runUMAP)
 export(run_cytonorm)
 export(runtSNE)
 export(scaleColors)
+export(subsample_geosketch)
 export(subset_fcd)
 export(subset_fcd_byparam)
 export(train_classifier_model)
 export(train_cytonorm)
 export(train_transfer_model)
 export(transform_data)
 export(used_markers)
+export(write_fcs)
 import(Biobase)
 import(CytoDx)
 import(CytoML)
@@ -74,6 +79,7 @@ import(Rtsne)
 import(SingleCellExperiment)
 import(SummarizedExperiment)
 import(caret)
+import(corrplot)
 import(cowplot)
 import(destiny)
 import(devtools)
@@ -101,7 +107,10 @@ import(stringr)
 import(tidyr)
 import(umap)
 import(uwot)
+importFrom(Biobase,AnnotatedDataFrame)
 importFrom(Matrix,sparseMatrix)
+importFrom(grDevices,recordPlot)
 importFrom(igraph,membership)
 importFrom(utils,packageDescription)
+importFrom(utils,sessionInfo)
 importFrom(utils,write.csv)

NAMESPACE

@@ -1,3 +1,17 @@
+# cyCONDOR 0.3.0
+* Bug fix with data projection where specific parameters were removed in UMAP calculation
+* Updated official Docker image to the latest Bioconductor version (Bioconductor 3.20, R 4.4.2)
+* Completed documentation of the `Astir` workflow.
+* Added parameter to select number of PCs and markers for UMAP projection in `learnUMAP` function.
+* Added documentation and guide to install cyCONDOR in Silicon Mac. We now how a compiled version of the Rphenoannoy package.
+* Simplified installation process, now automatically installing also the GitHub dependencies.
+* Enable Geometric Sketching to subset a `condor` object with the function `geometric_sketching`
+* Developed a function (`condor_session_info`) to include the `SessionInfo` to the `condor` object to increase tackability of the analysis.
+* Included storage of the statistical test results in the `condor` object
+* Added the option to display statistics to the `plot_frequency_boxplot` function - so far for `wilcox`, `t_test` and `diffcyt`
+* Added function to correlate the cyCONDOR results and the manual gating with FlowJo or similar software (`corr_plot_comparison`)
+* **Experimental**: Added a function to write FCS files from a `condor` object (`write_fcs`). Currently the function if writing in the .fcs file the cyCONDOR transformed values, we will change to the original values in the next iteration after more extensive testing.
+
 # cyCONDOR 0.2.1
 * Included help function to assign metaclusters (Thanks to Lucas Secchim Ribeiro)
 * Removal of redundant `color_palette` parameter and other redundant lines in the source code

NEWS.md

@@ -0,0 +1,78 @@
+# This function is designed to calculate a sample-wise correlation of cell population percentages that are found in CyCondor and FlowJo.
+
+
+#' corr_plot_comparison
+#' @title Sample-wise correlation of cell type proportions to compare manual gating and cyCONDOR.
+#' @description 'corr_plot_comparison' performs a sample-wise correlation of cell type proportions obtained via manual gating and via clustering on cyCONDOR. The result is shown in a correlogram.
+#' @param condor_df data frame containing the cell type frequencies obtained via clustering and annotation in cyCONDOR.
+#' @param flowjo_df data frame containing the cell type frequencies obtained via manual gating and annotation (e.g. FlowJo).
+#' @param sample_col name of the column containing sample names. This column name needs to be matching between the two data frames.
+#' @param method_corr correlation method used in the correlogram (default = "pearson").
+#' @param tl.cex Numeric, for the size of text label (variable names).
+#' @param cl.cex Numeric, text size of number-label in color-legend
+#' @import corrplot
+#' @importFrom grDevices recordPlot
+#'
+#' @export
+
+
+
+corr_plot_comparison <- function(condor_df,
+                                 flowjo_df,
+                                 sample_col,
+                                 method_corr = "pearson",
+                                 tl.cex = 1,
+                                 cl.cex = 1){
+
+
+     # Merge the two dataframes by Sample/DONOR ID
+     merged_data <- merge(condor_df, flowjo_df, by = sample_col)
+
+     flowjo_cols <- colnames(merged_data)[grepl("FlowJo",colnames(merged_data))]
+     condor_cols <- colnames(merged_data)[grepl("Condor",colnames(merged_data))]
+
+
+     condor_clean <- setNames(condor_cols, sub("^[^_]+_", "", condor_cols))
+     flowjo_clean <- setNames(flowjo_cols, sub("^[^_]+_", "", flowjo_cols))
+
+
+     matching_cell_types <- intersect(names(condor_clean), names(flowjo_clean))
+
+     total_cell_types_count <- length(condor_clean)
+     matching_cell_types_count <- length(matching_cell_types)
+
+
+
+     # Check if there are cell types not matching
+     diff_count <- total_cell_types_count - matching_cell_types_count
+
+     if (diff_count != 0) {
+
+       missing_cell_types <- setdiff(names(condor_clean), matching_cell_types)
+
+       warning(paste("There are ", diff_count, " cell types that are not matching between Condor and FlowJo.\n",
+                      "Missing cell types: ", paste(missing_cell_types,collapse = ", ")))
+
+       if(diff_count == total_cell_types_count){
+
+         stop("No matching cell types found. Ensure column names are formatted correctly and match between Condor and FlowJo.")
+       }
+     }
+
+
+     # Subset the two data frames only for the matching cell types
+     selected_condor <- condor_clean[matching_cell_types]
+     selected_flowjo <- flowjo_clean[matching_cell_types]
+
+     # Create a dataframe with only the selected columns
+     cor_data <- merged_data[, c(selected_condor, selected_flowjo)]
+
+
+     M <- cor(cor_data,use = "pairwise.complete.obs",method = method_corr)
+
+     corr_M <- M[selected_condor,selected_flowjo]
+
+     plot <- corrplot::corrplot(corr_M,method = "circle",type = "lower", tl.cex = 1, cl.cex = 1)
+
+     return(plot)
+}

R/corr_plot_comparison.R

@@ -522,3 +522,125 @@ prep_fjw <- function(data_gs,
 
 }
 
+
+#' Save the results from cyCONDOR as FCS file(s)
+#'
+#' @title Save the results from cyCONDOR as FCS file(s)
+#' @description Saves the expression data, annotation data and (optionally) results from the cyCONDOR analysis (dimensionality reduction, clustering) as one or more FCS file(s).
+#' @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 split_by NULL or character string specifying a metadata variable (e.g. sample_ID) to split the data. If NULL, only one FCS file with all data is generated.
+#' @param dir string specifying the directory where the FCS file(s) is/are saved. Current working directory by default.
+#' @param filename string specifying the filename for the FCS file. If split_by is defined, an automatic suffix according to the defined variable is assigned to the FCS files.
+#' @importFrom Biobase AnnotatedDataFrame
+#' @import flowCore
+#' @return Message that FCS file(s) are saved in the defined directory.
+#'
+#' @export
+write_fcs <- function(fcd = condor,
+                      expr_slot = "orig",
+                      reduction_method = NULL,
+                      reduction_slot = NULL,
+                      cluster_slot = NULL,
+                      cluster_var = NULL,
+                      split_by = NULL,
+                      dir = paste0(getwd(), "/"),
+                      filename){
+
+  # Print a working to let people know this is an experimental functio
+  warning("This function is still experimental. Feel free to test if and report your experience, this function
+  is currently writing cyCONDOR transformed value, future revision will export original values.")
+
+  # Get the expression data
+  xdata <- fcd[["expr"]][[expr_slot]]
+  exp <- xdata
+
+  # Add dimensionl reduction information
+  if(!is.null(reduction_method)){
+    rdata <- fcd[[reduction_method]][[reduction_slot]]
+    exp <- cbind(exp, rdata)
+  }
+
+  # Add cluster labels (must be numeric!)
+  if(!is.null(cluster_slot)){
+    cluster <- fcd[["clustering"]][[cluster_slot]][[cluster_var]]
+    exp <- cbind(exp, cluster)
+  }
+
+  # Ensure all columns in exp are numeric
+  exp[] <- lapply(exp, function(x){
+    if (is.numeric(x)) return(x)
+    xnum <- as.numeric(as.character(x))
+    if (any(is.na(xnum) & !is.na(x)))
+      warning("Some values in column could not be converted to numeric.")
+    return(xnum)
+  })
+
+  # Hanlde splitting
+  if(!is.null(split_by)){
+    anno <- fcd[["anno"]][["cell_anno"]]
+
+    if(!split_by %in% colnames(anno)){
+      stop("Column '", split_by, "' not found in the annotation data.")
+    }
+
+    split_vec <- anno[[split_by]]
+    exp_mat <- as.matrix(exp)
+    split_list <- split(seq_len(nrow(exp)), split_vec)
+
+    # create a flowSet with flowFrames for each subset
+    ffs <- lapply(names(split_list), function(group){
+      idx <- split_list[[group]]
+      exp_sub <- exp_mat[idx, , drop = FALSE]
+
+      # Create AnnotatedDataFrame for parameters of each subset
+      param_sub <- Biobase::AnnotatedDataFrame(data.frame(
+        name = colnames(exp_sub),
+        desc = paste("Channel", colnames(exp_sub), sep = "_"),
+        range = apply(exp_sub, 2, function(x) ceiling(max(x))),
+        minRange = apply(exp_sub, 2, min),
+        maxRange = apply(exp_sub, 2, max)
+      ))
+
+      # flowFrames
+      flowCore::flowFrame(exprs = exp_sub, parameters = param_sub)
+    })
+
+    names(ffs) <- names(split_list)
+
+    # flowSet
+    fs <- flowCore::flowSet(ffs)
+
+    # Write each frame into its own fcs file
+    for(sample_name in flowCore::sampleNames(fs)){
+      out_file <- paste0(dir, filename, "_", make.names(sample_name), ".fcs")
+      flowCore::write.FCS(fs[[sample_name]], file = out_file)
+      message("File was saved to ", out_file)
+    }
+  }
+
+  # No splitting, yields a single fcs file
+  else{
+
+    # Create AnnotatedDataFrame for parameters
+    param <- Biobase::AnnotatedDataFrame(data.frame(
+      name = colnames(exp),
+      desc = paste("Channel", colnames(exp), sep = "_"),
+      range = apply(exp, 2, function(x) ceiling(max(x))),
+      minRange = apply(exp, 2, min),
+      maxRange = apply(exp, 2, max)
+    ))
+
+    # Create flowFrame file
+    ffs <- flowCore::flowFrame(exprs = as.matrix(exp),
+                               parameters = param)
+
+    # Write the fcs
+    flowCore::write.FCS(ffs, file = paste0(dir, filename, ".fcs"))
+    message("File was saved to ", dir, filename, ".fcs")
+  }
+}