Merge pull request #13 from federicomarini/mosdefied-ideal

Mosdefied ideal

.Rbuildignore

@@ -9,5 +9,6 @@ _pkgdown.yml
 docs/*
 ^\.github/workflows/R-CMD-check\.yaml$
 ^\.github/workflows/pr-commands\.yaml$
+^\.github/workflows$
 ^CODE_OF_CONDUCT\.md$
 ^LICENSE\.md$

DESCRIPTION

@@ -1,30 +1,31 @@
 Package: ideal
 Type: Package
 Title: Interactive Differential Expression AnaLysis
-Version: 1.29.0
-Date: 2024-04-07
+Version: 1.99.0
+Date: 2024-09-13
 Authors@R: c(person("Federico", "Marini", role = c("aut", "cre"), 
     email ="[email protected]", 
     comment = c(ORCID = '0000-0003-3252-7758')))
 Description: This package provides functions for an Interactive
     Differential Expression AnaLysis of RNA-sequencing datasets, to
     extract quickly and effectively information downstream the step
     of differential expression. A Shiny application encapsulates
-    the whole package.
+    the whole package. Support for reproducibility of the whole analysis is 
+    provided by means of a template report which gets automatically compiled 
+    and can be stored/shared.
 License: MIT + file LICENSE
-LazyData: TRUE
 Depends: topGO
 Imports: 
     DESeq2, 
     SummarizedExperiment, 
+    mosdef (>= 1.1.0),
     GenomicRanges, 
     IRanges,
     S4Vectors, 
     ggplot2 (>= 2.0.0), 
     heatmaply, 
     plotly,
     pheatmap,
-    pcaExplorer, 
     IHW, 
     gplots, 
     UpSetR, 
@@ -48,8 +49,11 @@ Imports:
     shinyAce, 
     BiocParallel, 
     grDevices, 
+    graphics,
     base64enc,
-    methods
+    methods,
+    utils,
+    stats
 Suggests: 
     testthat, 
     BiocStyle, 
@@ -58,6 +62,7 @@ Suggests:
     org.Hs.eg.db,
     TxDb.Hsapiens.UCSC.hg38.knownGene, 
     DEFormats, 
+    htmltools,
     edgeR
 URL: https://github.com/federicomarini/ideal,
     https://federicomarini.github.io/ideal/
@@ -67,3 +72,4 @@ biocViews: ImmunoOncology, GeneExpression, DifferentialExpression, RNASeq, Seque
 VignetteBuilder: knitr
 RoxygenNote: 7.3.2
 Encoding: UTF-8
+Roxygen: list(markdown = TRUE)

NAMESPACE

@@ -19,7 +19,6 @@ import(SummarizedExperiment)
 import(ggplot2)
 import(knitr)
 import(methods)
-import(pcaExplorer)
 import(rmarkdown)
 import(shiny)
 import(shinydashboard)
@@ -47,11 +46,26 @@ importFrom(goseq,goseq)
 importFrom(goseq,nullp)
 importFrom(gplots,venn)
 importFrom(grDevices,axisTicks)
+importFrom(grDevices,colorRamp)
 importFrom(grDevices,dev.off)
 importFrom(grDevices,pdf)
+importFrom(grDevices,rgb)
+importFrom(graphics,pairs)
+importFrom(graphics,par)
+importFrom(graphics,strwidth)
+importFrom(graphics,text)
 importFrom(heatmaply,heatmaply)
 importFrom(limma,goana)
 importFrom(limma,topGO)
+importFrom(mosdef,create_link_ENSEMBL)
+importFrom(mosdef,create_link_GO)
+importFrom(mosdef,create_link_NCBI)
+importFrom(mosdef,deresult_to_df)
+importFrom(mosdef,gene_plot)
+importFrom(mosdef,get_annotation_orgdb)
+importFrom(mosdef,pair_corr)
+importFrom(mosdef,run_goseq)
+importFrom(mosdef,run_topGO)
 importFrom(pheatmap,pheatmap)
 importFrom(plotly,plotlyOutput)
 importFrom(plotly,renderPlotly)
@@ -68,6 +82,23 @@ importFrom(shinyAce,updateAceEditor)
 importFrom(shinyBS,bsCollapse)
 importFrom(shinyBS,bsCollapsePanel)
 importFrom(shinyBS,bsTooltip)
+importFrom(stats,as.formula)
+importFrom(stats,cor)
+importFrom(stats,quantile)
+importFrom(stats,sd)
+importFrom(stats,terms.formula)
+importFrom(stats,var)
 importFrom(stringr,str_count)
+importFrom(utils,count.fields)
+importFrom(utils,data)
+importFrom(utils,head)
+importFrom(utils,installed.packages)
+importFrom(utils,packageDescription)
+importFrom(utils,packageVersion)
+importFrom(utils,read.delim)
+importFrom(utils,sessionInfo)
+importFrom(utils,str)
+importFrom(utils,tail)
+importFrom(utils,write.csv)
 importMethodsFrom(GOstats,hyperGTest)
 importMethodsFrom(GOstats,summary)

NEWS.md

@@ -1,3 +1,14 @@
+# ideal 1.99.0
+
+## Other notes
+
+* The transition to the functions available in the `mosdef` Bioconductor is complete, with the original functions now being deprecated. This applies to `goseqTable()` (now replaced by `mosdef::run_goseq()`), which has now been made faster and more robust in its functionality and in the ways it can be executed
+* The gene plot widgets now also use the `gene_plot()` function from `mosdef`, instead of the previous `ggplotCounts()` function - `gene_plot()` is more flexible and has more options to control the behavior of the final plot object
+* The `deseqresult2tbl()` and `deseqresult2DEgenes()` are now replaced by the more flexible `mosdef::deresult_to_df()`
+* The internally defined `createLinkENS()`, `createLinkGeneSymbol()`, and `createLinkGO()` are now replaced by the equivalent functions in `mosdef`
+* The Roxygen-based documentation now supports markdown. No visible changes should appear to the user, as the content should have stayed fairly the same
+* Although no visible changes for the end user are expected, the incoming major version bump will reflect the change in the dependency graph, ensuring that this is noticed at least at the version numbering level
+
 # ideal 1.18.0
 
 ## Bug fixes

R/deprecated.R

@@ -0,0 +1,45 @@
+#' Deprecated functions in ideal
+#'
+#' Functions that are on their way to the function afterlife.
+#' Their successors are also listed.
+#' 
+#' The successors of these functions are likely coming after the rework that
+#' led to the creation of the `mosdef` package. See more into its 
+#' documentation for more details.
+#' 
+#' @param ... Ignored arguments.
+#' 
+#' @return All functions throw a warning, with a deprecation message pointing 
+#' towards its descendent (if available).
+#' 
+#' @name deprecated
+#' 
+#' @section Transitioning to the mosdef framework:
+#' 
+#' - [goseqTable()] is now being replaced by the more flexible 
+#' [mosdef::run_goseq()] function (which is even faster)
+#' - [ggplotCounts()] is now being replaced by the more flexible, better 
+#' designed, and actually even more good looking [mosdef::gene_plot()] 
+#' function, with better default behavior and all.
+#' - [deseqresult2tbl()] and [deseqresult2DEgenes()] are now replaced by the 
+#' more flexible [mosdef::deresult_to_df()]
+#' - The internally defined functions `createLinkENS()`, `createLinkGeneSymbol()`, 
+#' and `createLinkGO()` are now replaced by the equivalent functions in `mosdef`: 
+#' [mosdef::create_link_ENSEMBL()], [mosdef::create_link_NCBI()] and 
+#' [mosdef::create_link_GO()]. Notably, the `mosdef` package expanded on the 
+#' concept of automatically generated buttons, taking this to the extreme of 
+#' efficiency with the [mosdef::buttonifier()] function
+#' 
+#' @author Federico Marini
+#' 
+#' @examples
+#' # try(goseqtable())
+#' 
+NULL
+
+
+## #' @export
+## #' @rdname defunct
+## trendVar <- function(...) {
+##   .Defunct("fitTrendVar")
+## }

R/genesignatures.R

@@ -12,7 +12,7 @@
 #' # this example reads in the freely available pathways from wikipathways
 #' \dontrun{
 #' mysigs <- read_gmt(
-#'   "http://data.wikipathways.org/20180910/gmt/wikipathways-20180910-gmt-Homo_sapiens.gmt"
+#'   "http://data.wikipathways.org/20240910/gmt/wikipathways-20240910-gmt-Homo_sapiens.gmt"
 #' )
 #' head(mysigs)
 #' # see how the gene identifiers are encoded as ENTREZ id
@@ -34,11 +34,11 @@ read_gmt <- function(gmtfile) {
 #' Plot a heatmap for the selected gene signature on the provided data, with the
 #' possibility to compactly display also DE only genes
 #'
-#' @param vst_data A \code{\link{DESeqTransform}} object - usually the variance
+#' @param vst_data A [DESeqTransform()] object - usually the variance
 #' stabilized transformed data, which will be used to extract the expression values
 #' @param my_signature A character vector, usually named, containing the genes
 #' which compose the gene signature
-#' @param res_data A \code{\link{DESeqResults}} object. If not provided, it can
+#' @param res_data A [DESeqResults()] object. If not provided, it can
 #' be computed during the execution of the application
 #' @param FDR Numeric value between 0 and 1, the False Discovery Rate
 #' @param de_only Logical, whether to display only DE genes belonging to the pathway -
@@ -57,13 +57,13 @@ read_gmt <- function(gmtfile) {
 #' @param scale_row Logical, whether to perform row-based standardization of the
 #' expression values
 #'
-#' @return A plot based on the \code{pheatmap} function
+#' @return A plot based on the `pheatmap` function
 #' @export
 #'
 #' @examples
 #' # with the well known airway package...
-#' library(airway)
-#' data(airway)
+#' library("airway")
+#' data("airway", package = "airway")
 #' airway
 #' dds_airway <- DESeq2::DESeqDataSetFromMatrix(assay(airway),
 #'   colData = colData(airway),
@@ -117,7 +117,7 @@ sig_heatmap <- function(vst_data, my_signature,
   }
 
   if (de_only) {
-    de_res <- deseqresult2DEgenes(res_data, FDR)
+    de_res <- mosdef::deresult_to_df(res_data, FDR)
     de_genes <- de_res$id
     de_to_keep <- rownames(mydata_sig) %in% de_genes
     mydata_sig <- mydata_sig[de_to_keep, ]

R/ggplotCounts.R

@@ -3,27 +3,30 @@
 #' Plot for normalized counts of a single gene, with jittered points superimposed
 #' on the boxplot
 #'
-#' Note: this function relies on the \code{\link{plotCounts}} function of DESeq2,
+#' Note: this function relies on the [plotCounts()] function of DESeq2,
 #' therefore pseudocounts of 0.5 are added to each point
 #'
-#' @param dds A \code{\link{DESeqDataSet}} object.
+#' @param dds A [DESeqDataSet()] object.
 #' @param gene A character, specifying the name of the gene to plot
 #' @param intgroup Interesting groups: a character vector of
-#' names in \code{colData(dds)} to use for grouping
-#' @param annotation_obj A \code{data.frame} object, with \code{row.names} as gene
-#' identifiers (e.g. ENSEMBL ids) and a column, \code{gene_name}, containing
+#' names in `colData(dds)` to use for grouping
+#' @param annotation_obj A `data.frame` object, with `row.names` as gene
+#' identifiers (e.g. ENSEMBL ids) and a column, `gene_name`, containing
 #' e.g. HGNC-based gene symbols. Optional.
 #' @param transform Logical value, corresponding whether to have log scale y-axis
 #' or not. Defaults to TRUE.
-#' @param labels_repel Logical value. Whether to use \code{ggrepel}'s functions to
+#' @param labels_repel Logical value. Whether to use `ggrepel`'s functions to
 #' place labels; defaults to TRUE.
 #'
-#' @return An object created by \code{ggplot}
+#' @return An object created by `ggplot`
+#' 
+#' @importFrom mosdef gene_plot
+#' 
 #' @export
 #'
 #' @examples
-#' library(airway)
-#' data(airway)
+#' library("airway")
+#' data("airway", package = "airway")
 #' airway
 #' dds_airway <- DESeq2::DESeqDataSetFromMatrix(assay(airway),
 #'   colData = colData(airway),
@@ -35,59 +38,21 @@
 #' )
 ggplotCounts <- function(dds, gene, intgroup = "condition", annotation_obj = NULL,
                          transform = TRUE, labels_repel = TRUE) {
-  df <- plotCounts(dds, gene, intgroup, returnData = TRUE)
-  df$sampleID <- rownames(df)
-
-  if (!is.null(annotation_obj)) {
-    genesymbol <- annotation_obj$gene_name[match(gene, annotation_obj$gene_id)]
-  } else {
-    genesymbol <- ""
-  }
-
-  jittered_df <- df
-  # jittered_df$conditionj <- jitter(as.numeric(factor(jittered_df$condition)))
-  jittered_df$countj <- jitter(jittered_df$count)
-
-  onlyfactors <- df[, match(intgroup, colnames(df))]
-  df$plotby <- interaction(onlyfactors)
-
-  # base_breaks <- function(n = 10){
-  #   function(x) {
-  #     axisTicks(log10(range(x, na.rm = TRUE)), log = TRUE, nint = n)
-  #   }
-  # }
-
-  p <-
-    ggplot(df, aes_string(x = "plotby", y = "count", col = "plotby")) +
-    geom_boxplot(outlier.shape = NA) +
-    # geom_text(data = jittered_df,aes(x=conditionj,y=countj,label=sampleID)) +
-    scale_x_discrete(name = "") +
-    geom_jitter(aes_string(x = "plotby", y = "count"),
-      position = position_jitter(width = 0.1)
-    ) +
-    scale_color_discrete(name = "Experimental\nconditions")
-
-  if (labels_repel) {
-    p <- p + ggrepel::geom_text_repel(aes_string(label = "sampleID"))
-  } else {
-    p <- p + geom_text(aes_string(label = "sampleID"), hjust = -.1, vjust = 0)
-  }
-
-  if (transform) {
-    p <- p + scale_y_log10(name = "Normalized counts (log10 scale)")
-  } else {
-    p <- p + scale_y_continuous(name = "Normalized counts")
-  }
-
-  # scale_y_log10(name="Normalized counts - log10 scale") +
-  # coord_cartesian(ylim = c())# ,limits=c(0.1,NA)) +
-  p <- p + theme_bw()
-
-  if (!is.null(annotation_obj)) {
-    p <- p + labs(title = paste0("Normalized counts for ", genesymbol, " - ", gene))
-  } else {
-    p <- p + labs(title = paste0("Normalized counts for ", gene))
-  }
-
-  p
+  .Deprecated(old = "ggplotCounts", new = "mosdef::gene_plot", 
+              msg = paste0(
+                "Please use `mosdef::gene_plot()` in replacement of the `ggplotCounts()` function, ",
+                "originally located in the ideal package. \nCheck the manual page for ",
+                "`?mosdef::gene_plot()` to see the details on how to use it, e.g. ",
+                "refer to the new parameter definition and naming"))
+
+  p <- mosdef::gene_plot(
+    de_container = dds,
+    gene = gene,
+    intgroup = intgroup,
+    annotation_obj = annotation_obj,
+    transform = transform,
+    labels_repel = labels_repel
+  )
+
+  return(p)
 }