Remove LFQDataWriter, lfq_write_table, and unused dependency imports

- Delete R/LFQDataWriter.R and R/tidyMS_write.R (removed writexl, DT,
  htmlwidgets, readr dependencies)
- Remove conflicted::conflict_prefer usage from vignettes, use
  library(dplyr) instead
- Replace GSEA code section in Comparing2Groups vignette with
  descriptive text and links to clusterProfiler documentation
- Clean up unused @importFrom directives across R source files
- Regenerate NAMESPACE and man/ pages

Author: wolski

NAMESPACE

@@ -20,7 +20,6 @@ export(LFQDataStats)
 export(LFQDataSummariser)
 export(LFQDataToSummarizedExperiment)
 export(LFQDataTransformer)
-export(LFQDataWriter)
 export(LR_test)
 export(MissingHelpers)
 export(Model)
@@ -77,7 +76,6 @@ export(level_specific_contrasts)
 export(lfq_power_t_test_proteins)
 export(lfq_power_t_test_quantiles)
 export(lfq_power_t_test_quantiles_V2)
-export(lfq_write_table)
 export(linfct_all_possible_contrasts)
 export(linfct_factors_contrasts)
 export(linfct_from_model)
@@ -196,7 +194,6 @@ import(dplyr)
 import(ggplot2)
 importFrom(MASS,rlm)
 importFrom(UpSetR,upset)
-importFrom(conflicted,conflict_prefer)
 importFrom(dplyr,across)
 importFrom(dplyr,all_of)
 importFrom(dplyr,anti_join)
@@ -246,7 +243,6 @@ importFrom(graphics,plot)
 importFrom(graphics,rect)
 importFrom(graphics,text)
 importFrom(gridExtra,grid.arrange)
-importFrom(htmlwidgets,saveWidget)
 importFrom(methods,is)
 importFrom(pheatmap,pheatmap)
 importFrom(plotly,ggplotly)

R/A_dataset_docu.R

@@ -1,4 +1,3 @@
-#' @importFrom conflicted conflict_prefer
 #' @importFrom forcats fct_relevel
 #' @importFrom dplyr across all_of anti_join arrange bind_cols bind_rows case_when count desc distinct filter group_by
 #' @importFrom dplyr inner_join left_join mutate nest_by
@@ -13,7 +12,6 @@
 #' @importFrom graphics par rect hist
 #' @importFrom graphics abline pairs plot text
 #' @importFrom grDevices colorRampPalette dev.off pdf png rainbow
-#' @importFrom htmlwidgets saveWidget
 #' @importFrom pheatmap pheatmap
 #' @importFrom plotly ggplotly
 #' @importFrom purrr map map2 map2_dbl map_lgl map_chr map_dbl reduce map_if map_dfc map_int map_df

R/AnalysisConfiguration.R

@@ -468,7 +468,6 @@ setup_analysis <- function(data, configuration, cc = TRUE,  from_factors = FALSE
 #'
 #'
 #' @export
-#' @seealso \code{\link{lfq_write_table}}
 #' @param data data.frame
 #' @param config AnlalysisConfiguration
 #' @family configuration

R/ContrastsPlotter.R

@@ -218,45 +218,9 @@ ContrastsPlotter <- R6::R6Class(
     #' @param rank default FALSE, if TRUE then rank of avgAbd is used.
     #' @return ggplot
     ma_plot = function(fc, colour, legend = TRUE, rank = TRUE){
-      if ( missing(fc))
-        fc <- self$fcthresh
-      if (missing(colour)) {
-        colour <- self$modelName
-      }
-      contrastDF <- self$contrastDF
-      if (!is.null(contrastDF[[self$avg.abundance]])) {
-        # pdf version
-        if (rank) {
-          rankcol <- paste0("rank_", self$avg.abundance)
-
-          contrastDF <- contrastDF |>
-            dplyr::group_by(!!sym(self$contrast)) |>
-            mutate(!!rankcol := rank(!!sym(self$avg.abundance)))
-
-          #contrastDF[[ rankcol ]] <- rank(contrastDF[[self$avg.abundance]])
-          fig <- private$.ma_plot(
-            contrastDF,
-            rankcol,
-            self$diff,
-            self$contrast,
-            fc,
-            colour = colour,
-            legend = legend )
-        }else{
-          fig <- private$.ma_plot(
-            contrastDF,
-            self$avg.abundance,
-            self$diff,
-            self$contrast,
-            fc,
-            colour = colour,
-            legend = legend)
-        }
-      }else{
-        warning("no group_1 group_2 columns can't generate MA")
-        fig <- NULL
-      }
-      return(fig)
+      if (missing(fc)) fc <- self$fcthresh
+      if (missing(colour)) colour <- self$modelName
+      private$.ma_fig(self$contrastDF, fc, colour, legend, rank)
     },
     #' @description
     #' ma plotly
@@ -266,49 +230,11 @@ ContrastsPlotter <- R6::R6Class(
     #' @param rank default FALSE, if TRUE then rank of avgAbd is used.
     #' @return list of ggplots
     ma_plotly = function(fc, colour, legend = TRUE, rank = FALSE){
-      # html version
-      if (missing(fc))
-        fc <- self$fcthresh
-      if (missing(colour))
-        colour <- self$modelName
-      contrastDF  <- self$contrastDF
-      if (!is.null(contrastDF[[self$avg.abundance]])) {
-        if (rank) {
-          rankcol <- paste0("rank_", self$avg.abundance)
-          contrastDF <- contrastDF |>
-            dplyr::group_by(!!sym(self$contrast)) |>
-            mutate(!!rankcol := rank(!!sym(self$avg.abundance)))
-
-          fig <- private$.ma_plot(
-            contrastDF,
-            rankcol,
-            self$diff,
-            self$contrast,
-            fc,
-            colour = colour,
-            legend = legend
-          )
-        } else {
-          contrastDF  <- contrastDF |>
-            plotly::highlight_key(~subject_Id)
-          fig <- private$.ma_plot(
-            contrastDF,
-            self$avg.abundance,
-            self$diff,
-            self$contrast,
-            fc,
-            colour = colour,
-            legend = legend
-          )
-        }
-
-        fig_plotly <- fig |>
-          plotly::ggplotly(tooltip = "subject_Id")
-
-        return(fig_plotly)
-      }else{
-        return(NULL)
-      }
+      if (missing(fc)) fc <- self$fcthresh
+      if (missing(colour)) colour <- self$modelName
+      fig <- private$.ma_fig(self$contrastDF, fc, colour, legend, rank, plotly_mode = TRUE)
+      if (!is.null(fig)) fig <- fig |> plotly::ggplotly(tooltip = "subject_Id")
+      fig
     },
     #' @description
     #' plot a score against the log2 fc e.g. t-statistic
@@ -383,6 +309,23 @@ ContrastsPlotter <- R6::R6Class(
     }
   ),
   private = list(
+    .ma_fig = function(contrastDF, fc, colour, legend, rank, plotly_mode = FALSE) {
+      if (is.null(contrastDF[[self$avg.abundance]])) {
+        if (!plotly_mode) warning("no group_1 group_2 columns can't generate MA")
+        return(NULL)
+      }
+      xcol <- self$avg.abundance
+      if (rank) {
+        xcol <- paste0("rank_", self$avg.abundance)
+        contrastDF <- contrastDF |>
+          dplyr::group_by(!!sym(self$contrast)) |>
+          dplyr::mutate(!!xcol := rank(!!sym(self$avg.abundance)))
+      } else if (plotly_mode) {
+        contrastDF <- contrastDF |> plotly::highlight_key(~subject_Id)
+      }
+      private$.ma_plot(contrastDF, xcol, self$diff, self$contrast, fc,
+                       colour = colour, legend = legend)
+    },
     .volcano = function(contrasts,
                         scores,
                         colour = NULL,

R/ContrastsProDA.R

@@ -115,16 +115,5 @@ ContrastsProDA <- R6::R6Class(
         columns = c("diff", columns),
         contrast = 'contrast')
       return(contrasts_wide)
-    },
-    #' @description write results
-    #' @param path directory
-    #' @param filename file to write to
-    #' @param format default xlsx \code{\link{lfq_write_table}}
-    write = function(path, filename, format = "xlsx"){
-      filename <- if (missing(filename)) {self$modelName} else (filename )
-      lfq_write_table(self$get_contrasts(),
-                      path = path,
-                      name  = paste0("Contrasts_",filename),
-                      format = format)
     }
   ))

R/LFQData.R

@@ -2,6 +2,13 @@
 #'
 #' LFQData R6 class
 #'
+#' @section Missing Data Assumptions:
+#' The filtering and imputation methods in this package assume that
+#' missing values are Missing Completely At Random (MCAR) or Missing At
+#' Random (MAR). Abundance-dependent missingness (MNAR), which is common
+#' in DDA proteomics, is not modelled. Users should be aware that MNAR
+#' can bias fold-change estimates and inflate false discovery rates.
+#'
 #' @export
 #' @family LFQData
 #' @examples

R/LFQDataImp.R

@@ -74,6 +74,11 @@ function_lod_quantile <- function(data_matrix, percent = 10) {
 #'   "GBM", "SQ", "BL", or "CZM" (passed to zCompositions)
 #' @param lod limit of detection strategy, either "global" or "quantile"
 #' @return the modified LFQData object (lfqdata), with imputed values
+#' @note This function assumes that missing values are Missing Completely
+#'   At Random (MCAR) or Missing At Random (MAR). If missingness is
+#'   abundance-dependent (MNAR, common in proteomics DDA), the imputed
+#'   values and downstream statistics may be biased. For MNAR-aware
+#'   analysis, consider packages such as proDA or DEqMS.
 #' @export
 #'
 #' @examples

R/LFQDataSummariser.R

@@ -101,23 +101,19 @@ LFQDataSummariser <- R6::R6Class(
     #' Computes the percent abundance of proteins overall and within each group
     #' @return data frame
     percentage_abundance = function(){
-      # roll up to protein intensities
-      # compute protein level summaries
-
       dall <- prolfqua::summarize_stats_all(self$lfq$data, self$lfq$config)
       dfac <- prolfqua::summarize_stats_factors(self$lfq$data, self$lfq$config)
-
       all <- dplyr::bind_rows(dfac, dall)
-      nested <- all |> dplyr::group_by(!!sym("interaction")) |> tidyr::nest()
-      for (i in seq_len(nrow(nested))) {
-        nested$data[[i]] <- nested$data[[i]] |>
-          dplyr::arrange(.data$meanAbundance) |>
-          dplyr::mutate(id = dplyr::row_number()) |>
-          dplyr::mutate(abundance_percent = meanAbundance/sum(meanAbundance, na.rm = TRUE)*100 ) |>
-          dplyr::mutate(abundance_percent_cumulative = cumsum(ifelse(is.na(abundance_percent), 0, abundance_percent)) + abundance_percent*0) |>
-          dplyr::mutate(percent_prot = id / max(id) * 100)
-      }
-      res <- tidyr::unnest(nested, cols = "data")
+      res <- all |>
+        dplyr::group_by(!!sym("interaction")) |>
+        dplyr::arrange(.data$meanAbundance, .by_group = TRUE) |>
+        dplyr::mutate(
+          id = dplyr::row_number(),
+          abundance_percent = meanAbundance / sum(meanAbundance, na.rm = TRUE) * 100,
+          abundance_percent_cumulative = cumsum(ifelse(is.na(abundance_percent), 0, abundance_percent)) + abundance_percent * 0,
+          percent_prot = id / max(id) * 100
+        ) |>
+        dplyr::ungroup()
       return(res)
     }
 

R/LFQDataWriter.R

@@ -109,16 +109,6 @@ Model <- R6::R6Class(
       return(dplyr::ungroup(Model_Anova))
     },
 
-    #' @description
-    #' writes model coefficients to file
-    #' @param path folder to write to
-    #' @param format default xlsx \code{\link{lfq_write_table}}
-    write_coefficients  = function(path, format = "xlsx"){
-      lfq_write_table(self$get_coefficients(),
-                      path = path,
-                      name  = paste0("Coef_",self$modelName),
-                      format = format)
-    },
     #' @description
     #' histogram of model coefficient
     coef_histogram = function(){