Merge pull request #4 from pfizer-opensource/copilot/add-dili-discovery-proteomics-vignette

Add DILI Discovery Proteomics vignette

Author: Rvirgenslane

BuildSupport/BuildMe.R

@@ -39,6 +39,37 @@ devtools::document(); devtools::load_all()
 rmarkdown::render("README.Rmd", 
                   output_file="README.md")
 
+rm(list = ls()); gc()
+devtools::document(); devtools::load_all()
+
+# build images
+source(file.path(getwd(), "data-raw/shiny_ui/shiny_shots_optimized.R"))
+
+
+list_rmds <- file.path(getwd(), "vignettes") |> 
+  list.files(pattern = ".Rmd", full.names = TRUE) |> 
+  stringr::str_subset("0[1-9]|DILI") 
+
+rendering_list_rmds <-list_rmds #[6]
+
+# file.edit(list_rmds[3])
+
+rendering_list_rmds|> 
+  purrr::set_names(~basename(.x)) |> 
+  purrr::iwalk(function(x,y){
+    
+    md_name <- y |> 
+      stringr::str_replace("[.]Rmd", ".md") 
+    
+rmarkdown::render(x, 
+output_dir = file.path(getwd(), "vignettes"),
+output_file = md_name)
+    
+  })
+
+
+
+
 
 devtools::test(stop_on_failure = TRUE)
 

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: Hotgenes
 Type: Package
 Title: Tools to simplify Omics DE Analysis
-Version: 0.0.54
+Version: 0.0.55
 Author: Richard Virgen-Slane <Richard.Virgen-Slane@pfizer.com>
 Maintainer: Richard Virgen-Slane <Richard.Virgen-Slane@pfizer.com>
 Description: Converts outputs from DESeq2, limma, or a method of your
@@ -13,11 +13,10 @@ Encoding: UTF-8
 LazyData: true
 RoxygenNote: 7.3.3
 Imports: 
-    BiocParallel,
     cli,
     factoextra,
     FactoMineR,
-    fgsea (>= 1.16.0),
+    fgsea (>= 1.32.0),
     forcats,
     DRomics,
     DT,
@@ -59,6 +58,7 @@ Suggests:
     airway,
     apeglm,
     ashr,
+    BiocParallel,
     edgeR,
     devtools,
     GEOquery,

NAMESPACE

@@ -114,7 +114,6 @@ export(wrap_fixed_names)
 exportMethods("Mapper_<-")
 exportMethods("auxiliary_assays_<-")
 exportMethods("coldata_<-")
-importFrom(BiocParallel,MulticoreParam)
 importFrom(DESeq2,lfcShrink)
 importFrom(DESeq2,normTransform)
 importFrom(DESeq2,resultsNames)

NEWS

@@ -1,3 +1,16 @@
+# version 0.0.55:
+    - Fixed contrast_vectors arg to HotgenesDEseq2() bug that dropped contrast
+      names from export.
+    - Fixed HotgenesDEseq2() issue not supporting stat col generation for "ashr"
+    - Updates to DE() to stabilize Topn behavoir across Report types.
+    - Reorganized the package vignettes into a clearer 01–05 documentation suite.
+    - Expanded the visualization vignette to include GSEA, msigdbr_wrapper(),
+      and custom gene-set configuration examples.
+    - Refocused the Shiny vignette on using the app to prioritize genes of
+      interest across DE, PCA, Venn, GSEA, and expression views.
+    - Added a separate vignette for sample-wise pathway activity with
+      HotgeneSets().
+
 # version 0.0.54:
     - Replaced base R messaging/error calls (stop(), warning(), message(),
       print(), cat()) with cli equivalents across R/ sources.

R/GSEA.R

@@ -22,7 +22,7 @@ fgsea_ <- function(
     pathways = NULL,
     minSize = 1,
     maxSize = Inf,
-    nproc = 1,
+    nproc = 0,
     verbose = FALSE,
     ...) {
   # Check if Ranks is a list

R/Hotgenes.R

@@ -199,6 +199,7 @@ HotgenesDEseq2 <- function(DEseq2_object = NULL,
 
   } else {
     output_method <- Output_contrasts %>%
+      purrr::set_names(~ .x) %>%
       purrr::imap(function(xList, yList) {
         list(
 
@@ -209,6 +210,9 @@ HotgenesDEseq2 <- function(DEseq2_object = NULL,
   }
 
 
+  DESeq2_args <- tryCatch(list(...), 
+                       error = function(e) NULL)
+  
   Output_DE <- output_method %>%
     purrr::set_names(~ .x) %>%
     purrr::imap(function(xList, yList) {
@@ -218,16 +222,17 @@ HotgenesDEseq2 <- function(DEseq2_object = NULL,
         dds = DEseq2_object
          ) |> 
         append(xList) |> 
-        append(list(...))
+        append(DESeq2_args)
 
       out_lfcSh <- base::do.call(what = DESeq2::lfcShrink,
                                 args = list_lfcSh ) |> 
         base::as.data.frame()  |> 
         tibble::rownames_to_column(var = "Feature")  |> 
         dplyr::arrange(.data$padj)
 
-      # Calculating stat for GSEA
-      if (lfcShrink_type == "apeglm") {
+      
+      # Calculating stat for GSEA when not returned by lfcShrink
+      if (!"stat" %in% names(out_lfcSh)) {
         out_lfcSh <- out_lfcSh %>%
           dplyr::mutate(stat = (-log10(.data$padj) * sign(.data$log2FoldChange)))
       }

R/Shiny_Hotgenes_UI_Module.R

@@ -209,7 +209,7 @@ return(UI_params)
 #' Mapper_ columns that should be used to map expression data
 #' to ontology gene identifiers. Default is names(Mapper_(Hotgenes)).
 #' A named vector may be used, as well.
-#' @param parallel.sz Number of threads of execution
+#' @param nproc Number of threads of execution
 #' to use when doing the calculations in parallel.
 #' @param remove_modal_spinner logical, if TRUE remove_modal_spinner will
 #' run at the end of initial server run.
@@ -225,7 +225,7 @@ session = NULL,
 Hotgenes = NULL,
 OntologyMethods = Hotgenes::OntologyMethods(),
 Mapper_choices = names(Mapper_(Hotgenes)),
-parallel.sz = 1L,
+nproc = 1L,
 max_col_levels = Inf,
 selected_fillby = "",
 remove_modal_spinner = TRUE) {
@@ -355,7 +355,7 @@ BoxPlot_output$reactive_NormSlot()
 # # verifies that everything is loaded
 if (isTRUE(remove_modal_spinner)) {
 shiny::observeEvent(shiny::req(serverOut() %in% ExpressionSlots_(Hotgenes_out)), {
-cli::cli_inform("remove_modal_spinner")
+
 shinybusy::remove_modal_spinner(session = session)
 })
 }
@@ -396,7 +396,7 @@ Hotgenes = NULL,
 OntologyMethods = Hotgenes::OntologyMethods(),
 Mapper_choices = names(Mapper_(Hotgenes)),
 theme = shinythemes::shinytheme("united"),
-parallel.sz = 1L,
+nproc = 1L,
 max_col_levels = Inf)
 UseMethod("Shiny_Hotgenes", Hotgenes)
 
@@ -409,7 +409,7 @@ Hotgenes = NULL,
 OntologyMethods = Hotgenes::OntologyMethods(),
 Mapper_choices = names(Mapper_(Hotgenes)),
 theme = shinythemes::shinytheme("united"),
-parallel.sz = 1L,
+nproc = 1L,
 max_col_levels = Inf) {
 
 
@@ -457,7 +457,7 @@ Hotgenes = NULL,
 OntologyMethods = Hotgenes::OntologyMethods(),
 Mapper_choices = names(Mapper_(Hotgenes)),
 theme = shinythemes::shinytheme("united"),
-parallel.sz = 1L,
+nproc = 1L,
 max_col_levels = Inf) {
 
 cli::cli_inform("processing as list")
@@ -568,7 +568,7 @@ session = session,
 OntologyMethods = OntologyMethods,
 max_col_levels = max_col_levels,
 Hotgenes = Hotgenes_out() ,
-parallel.sz = parallel.sz,
+nproc = nproc,
 id = "Obj_A"
 )
 

R/Shiny_Tabs.R

@@ -1685,13 +1685,12 @@ shiny::plotOutput("enrichmentPlot" %>% ns())
 #' @inheritParams DE_pheServer
 #' @inheritParams fgsea_
 #' @inheritParams Shiny_Hotgenes_Server
-#' @importFrom BiocParallel MulticoreParam
 #' @param OntologyMethods output of OntologyMethods function.
 #' @param Mapper_choices vector containing names of
 #' Mapper_ columns that should be used to map expression data
 #' to ontology gene identifiers. Default is names(Mapper_(Hotgenes)).
 #' A named vector may be used, as well.
-#' @param parallel.sz Number of threads of execution
+#' @param nproc Number of threads of execution
 #' to use when doing the calculations in parallel.
 #' @param PCA_TopTibble reactive object containing the TopTibble slots returned
 #' by PCA.
@@ -1708,7 +1707,7 @@ ExpressionSlots = shiny::reactive(NULL),
 SampleIDs = shiny::reactive(NULL),
 OntologyMethods = OntologyMethods(),
 Mapper_choices = names(Mapper_(Hotgenes)),
-parallel.sz = 1) {
+nproc = 0) {
 shiny::moduleServer(
 id,
 function(input, output, session) {
@@ -1812,10 +1811,12 @@ input$fgsea_Button
 
 # must be observed
 shiny::observe({
-paste0(
-"Setting GSEA ini_fgsea() = ",
-ini_fgsea()
-) 
+# this_message <- paste0(
+# "Setting GSEA ini_fgsea() = ",
+# ini_fgsea()
+# )
+
+  cli::cli_inform("Setting GSEA ini_fgsea() = {ini_fgsea()}")
 })
 
 
@@ -1906,27 +1907,18 @@ if(gs_check == 0) {
 shiny::req(gs_check > 0)
 
 
-shinybusy::update_modal_spinner(
-paste0(
-"Checking ",
-length(pthyways),
-" pathways"
-),
-session = session
-)
-
-
 OutPut_fgsea <- fgsea_(
 Ranks = list_Ranks,
 pathways = pthyways,
-BPPARAM = BiocParallel::MulticoreParam(workers = parallel.sz),
+nproc = nproc,
 minSize = input$minSize_Onto,
 maxSize = input$maxSize_Onto
 )
 
 
 shinybusy::remove_modal_spinner(session = session)
 
+
 return(OutPut_fgsea)
 })
 

R/Utilities.R

@@ -173,10 +173,11 @@ DE <- function(
       purrr::set_names(levels(Output_DE[[dplyr::group_vars(Output_DE)]])) %>%
       purrr::map_int(function(x) {
         x %>%
-          head(n = Topn) %>%
+         
           dplyr::pull(.data$Feature) %>%
           # to bypass issue with multiple mappings due to mapFeatures = TRUE
           unique() %>%
+          head(n = Topn) %>%
           length()
       })
     return(ReportOut)
@@ -189,10 +190,11 @@ DE <- function(
       purrr::set_names(levels(Output_DE[[dplyr::group_vars(Output_DE)]])) %>%
       purrr::map(function(x) {
         x %>%
-          head(n = Topn) %>%
           dplyr::pull(.data$Feature) %>%
+          # to bypass issue with multiple mappings due to mapFeatures = TRUE
+          unique() %>%
+          head(n = Topn) 
 
-          unique()
       })
     return(ReportOut)
   }
@@ -202,10 +204,11 @@ DE <- function(
 
       plyr::dlply("contrast_dir", identity) %>% 
       purrr::map(function(x) {
-        x %>%
-          head(n = Topn) %>%
+        x  %>%
           dplyr::pull(.data$Feature) %>%
-          unique()
+          # to bypass issue with multiple mappings due to mapFeatures = TRUE
+          unique() %>%
+          head(n = Topn)
       })
     return(ReportOut)
   }
@@ -218,10 +221,8 @@ DE <- function(
       purrr::set_names(levels(Output_DE[[dplyr::group_vars(Output_DE)]])) %>%
       purrr::map(function(x) {
         x <- x %>%
-          head(n = Topn) %>%
-         
           dplyr::select(dplyr::all_of(c(Rank_name_Final, "stat"))) %>%
-          
+          dplyr::distinct() %>%
           dplyr::filter(
             dplyr::if_any(
               .cols = dplyr::any_of(Rank_name_Final),
@@ -233,7 +234,9 @@ DE <- function(
           dplyr::ungroup() %>%
           dplyr::arrange(dplyr::desc(.data$stat)) %>%
           unique() %>%
-          tibble::deframe()
+          head(n = Topn) %>%
+          tibble::deframe() 
+          
 
 
 

README.Rmd

@@ -24,6 +24,34 @@ options(tibble.print_min = 5, tibble.print_max = 5)
 
 <br>
 
+> For a guided introduction to the package, start with the **[vignette suite](#vignettes)** below.
+> The remainder of this README contains extended worked examples and required project information.
+
+## Contents
+
+- [Vignettes](#vignettes)
+- [Installation](#installation-requires-r--420-however-r--430-is-recommended)
+- [License information](#license-information)
+- [Background](#background)
+- [Why are Pfizer sharing this?](#why-are-pfizer-sharing-this)
+- [What is the benefit of this work?](#what-is-the-benefit-of-this-work)
+- [How should I submit questions, queries and enhancements?](#how-should-i-submit-questions-queries-and-enhancements)
+- [How does Hotgenes work?](#how-does-hotgenes-work)
+- [What can you do with a Hotgenes object?](#what-can-you-do-with-a-hotgenes-object)
+- [All functionality is available in a shiny app!](#all-functionality-is-available-in-a-shiny-app)
+- [Explore functions](#explore-functions)
+- [Code of Conduct](#code-of-conduct)
+
+## Vignettes
+
+The package documentation is organized into the following vignettes:
+
+- [01 Creating Hotgenes Objects](vignettes/01_Creating_Hotgenes_Objects.md)
+- [02 API and Methods](vignettes/02_API_and_Methods.md)
+- [03 Visualization, Exploration, and Enrichment](vignettes/03_Visualization_and_Exploration.md)
+- [04 Using Shiny to Prioritize Genes of Interest](vignettes/04_Interactive_Exploration_Shiny.md)
+- [05 Sample-wise Pathway Activity with HotgeneSets](vignettes/05_Samplewise_Pathway_Activity_HotgeneSets.md)
+
 ## Installation requires R (>= 4.2.0); however R (>= 4.3.0) is recommended
 
 ```{r install_reqs, eval=FALSE, include=TRUE}