add child computation

NAMESPACE

@@ -102,7 +102,7 @@ export(names_to_matrix)
 export(normalize_log2_robscale)
 export(nr_B_in_A)
 export(nr_B_in_A_per_sample)
-export(nr_obs_hierarchy)
+export(nr_obs_experiment)
 export(nr_obs_sample)
 export(old2new)
 export(pairs_smooth)

R/AnalysisConfiguration.R

@@ -396,15 +396,16 @@ hierarchy_counts <- function(pdata, config){
 #'
 #' config <- bb$config
 #' data <- bb$data
-#' res <- hierarchy_counts_sample(data, config)
+#' res <- hierarchy_counts_sample(data, config, nr_children = 1)
 #' res()
 #' x <- res("long")
 #' res("plot")
-#'
+#' # filters on peptide level
 #' res <- hierarchy_counts_sample(data, config, nr_children = 2)
 #' res()
 #' x2 <- res("long")
 #' res("plot")
+#' # filters on protein level based on peptide count
 #' bb <- prolfqua::sim_lfq_data_protein_config()
 #' res <- hierarchy_counts_sample(bb$data, bb$config, nr_children = 2)
 #' x1 <- res()
@@ -414,15 +415,23 @@ hierarchy_counts <- function(pdata, config){
 #' x2 <- res()
 #' res("long")
 #' res("plot")
-#' dplyr::inner_join(x1,
-#'  x2,
-#'  by =  c("isotopeLabel","sampleName"),
-#'  suffix =c("2","1"))
+#' x1$nr_children <- 2
+#' x2$nr_children <- 1
+#' xl <- dplyr::bind_rows(x1, x2)
+#'
+#'
+#'
+#'
+#' xl$nr_children |> table()
+#' nudgeval <-  -mean(xl$protein_Id) * 0.05
+#' ggplot(xl, aes(x = sampleName, y = protein_Id, fill = as.character(nr_children)) ) +
+#'  geom_bar(stat = "identity", position = position_dodge())
 #'
 hierarchy_counts_sample <- function(
     pdata,
     configuration,
     nr_children = 1) {
+
   hierarchy <- configuration$table$hierarchy_keys()
   summary <- pdata |> dplyr::filter(!is.na(!!rlang::sym(configuration$table$get_response() )),
                                     !!rlang::sym(configuration$table$nr_children ) >= .env$nr_children) |>
@@ -440,7 +449,6 @@ hierarchy_counts_sample <- function(
       if (value == "long") {
         return(long)
       }else if (value == "plot" & nrow(long) > 0) {
-
         nudgeval <-  -mean(long$nr) * 0.05
         # TODO(WEW) check potential problem with sampleName
         ggplot2::ggplot(long, ggplot2::aes(x = !!rlang::sym(configuration$table$sampleName), y = .data$nr)) +

R/Contrasts.R

@@ -49,6 +49,8 @@ Contrasts <- R6::R6Class(
     contrast_result = NULL,
     #' @field global use a global linear function (determined by get_linfct)
     global = TRUE,
+    #' @field protein_annot holds protein annotation
+    protein_annot = NULL,
     #' @description
     #' initialize
     #' create Contrast

R/ContrastsPlotter.R

@@ -76,6 +76,8 @@ ContrastsPlotter <- R6::R6Class(
     fcthresh = 1,
     #' @field avg.abundance name of column containing avg abundance values.
     avg.abundance = character(),
+    #' @field avg.abundance name of column containing avg abundance values.
+    protein_annot = NULL,
     #' @description
     #' create Crontrast_Plotter
     #' @param contrastDF frame with contrast data
@@ -98,7 +100,8 @@ ContrastsPlotter <- R6::R6Class(
                           modelName = "modelName",
                           diff = "diff",
                           contrast = "contrast",
-                          avg.abundance = "avgAbd"
+                          avg.abundance = "avgAbd",
+                          protein_annot = NULL
     ){
       self$contrastDF <- tidyr::unite(
         contrastDF,
@@ -113,6 +116,7 @@ ContrastsPlotter <- R6::R6Class(
       self$fcthresh = fcthresh
       self$contrast = contrast
       self$avg.abundance = avg.abundance
+      self$protein_annot = protein_annot
     },
     #' @description
     #' plot histogram of selected scores (e.g. p-value, FDR, t-statistics)

R/ProteinAnnotation.R

@@ -6,16 +6,13 @@
 #' @examples
 #'
 #' istar <-prolfqua::sim_lfq_data_peptide_config()
-#' data <- istar$data
-#'
-#'
-#' lfqdata <- LFQData$new(data, istar$config)
+#' lfqdata <- LFQData$new(istar$data, istar$config)
 #' pannot <- ProteinAnnotation$new( lfqdata )
 #' pannot$annotate_decoys()
 #' pannot$annotate_contaminants()
 #' dd <- pannot$clean()
 #' tmp <- lfqdata$get_subset(dd)
-#'
+#' pannot$row_annot
 #'
 ProteinAnnotation <-
   R6::R6Class("ProteinAnnotation",
@@ -49,8 +46,12 @@ ProteinAnnotation <-
                   } else {
                     self$row_annot <- distinct(select(lfqdata$data, self$pID))
                   }
-
-
+                  if (!self$nr_peptides %in% colnames(row_annot) ) {
+                    self$row_annot <- inner_join(
+                      self$row_annot,
+                      nr_obs_experiment(lfqdata$data, lfqdata$config, name_nr_child = self$nr_peptides),
+                      by = self$pID)
+                  }
                 },
                 #' @description
                 #' annotate rev sequences

R/tidyMS_R6_TransitionCorrelations.R

@@ -709,28 +709,37 @@ nr_obs_sample <- function(data, config, new_child = config$table$nr_children){
 #' Aggregates e.g. protein abundances from peptide abundances
 #'
 #' @export
+#' @param data tidy data
+#' @param config prolfqua config
+#' @param from_children compute from existing child stats
+#' @param name_nr_child how to name column
 #' @examples
 #' dd <- prolfqua::sim_lfq_data_peptide_config()
 #'
-#' xd <- nr_obs_hierarchy(dd$data, dd$config)
+#' xd <- nr_obs_experiment(dd$data, dd$config)
+#' xd
+#' xd <- nr_obs_experiment(dd$data, dd$config, from_children = FALSE)
 #' xd
-#' dp <- prolfqua::sim_lfq_data_protein_config()
-#' debug(nr_obs_hierarchy)
-#' nr_obs_sample(dp$data, dp$config)
-#' xd <- nr_obs_hierarchy(dp$data, dp$config)
 #'
+#' dp <- prolfqua::sim_lfq_data_protein_config()
+#' undebug(nr_obs_experiment)
+#' nr_obs_experiment(dp$data, dp$config)
+#' nr_obs_experiment(dp$data, dp$config, from_children = FALSE)
 #'
-nr_obs_hierarchy <- function(data, config, from_children = TRUE , name_nr_child = "nr_child_exp"){
+nr_obs_experiment <- function(data, config, from_children = TRUE,
+                              name_nr_child = "nr_child_exp"){
   tb <- config$table
   if (!from_children & (tb$hierarchyDepth < length(tb$hierarchy_keys())) ) {
-    xq <- data |> tidyr::select(tb$hierarchy_keys()) |>
+    xq <- data |> dplyr::select(tb$hierarchy_keys()) |>
       distinct() |>
       dplyr::group_by(!!sym(tb$hierarchy_keys_depth())) |>
       dplyr::summarize(!!name_nr_child := dplyr::n(), .groups = "drop")
+    return(xq)
   } else {
     xz <- nr_obs_sample(data,config)
-    xz <- x |> group_by(!!sym(tb$hierarchy_keys_depth())) |>
+    xz <- xz |> group_by(!!sym(tb$hierarchy_keys_depth())) |>
       summarize(!!name_nr_child := max(!!sym(tb$nr_children)), .groups = "drop")
+    return(xz)
   }
 }