From c1150db64e8663e78ba0e904bba4be92e74414f9 Mon Sep 17 00:00:00 2001
From: Witold Wolski <wewolski@gmail.com>
Date: Sat, 6 Apr 2024 11:57:05 +0200
Subject: [PATCH] improve support for peptide counts

---
 NAMESPACE                            |  3 +-
 R/ProteinAnnotation.R                | 17 +++----
 R/tidyMS_R6_TransitionCorrelations.R | 66 ++++++++++++++++++++++++----
 R/tidyMS_aggregation.R               | 22 ++--------
 man/ProteinAnnotation.Rd             | 15 +++----
 man/get_imputed_contrasts.Rd         |  4 --
 man/nr_B_in_A_per_sample.Rd          | 13 ++----
 man/nr_obs.Rd                        | 20 ---------
 man/nr_obs_hierarchy.Rd              | 28 ++++++++++++
 man/nr_obs_sample.Rd                 | 24 ++++++++++
 10 files changed, 131 insertions(+), 81 deletions(-)
 delete mode 100644 man/nr_obs.Rd
 create mode 100644 man/nr_obs_hierarchy.Rd
 create mode 100644 man/nr_obs_sample.Rd

diff --git a/NAMESPACE b/NAMESPACE
index d058e9bf3..c317e6891 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -102,7 +102,8 @@ export(names_to_matrix)
 export(normalize_log2_robscale)
 export(nr_B_in_A)
 export(nr_B_in_A_per_sample)
-export(nr_obs)
+export(nr_obs_hierarchy)
+export(nr_obs_sample)
 export(old2new)
 export(pairs_smooth)
 export(pairs_w_abline)
diff --git a/R/ProteinAnnotation.R b/R/ProteinAnnotation.R
index 2058d94ee..237edac7a 100644
--- a/R/ProteinAnnotation.R
+++ b/R/ProteinAnnotation.R
@@ -5,22 +5,17 @@
 #' @family LFQData
 #' @examples
 #'
-#' istar <- prolfqua_data('data_ionstar')$filtered()
-#' istar$config <- old2new(istar$config)
-#' data <- istar$data #|> dplyr::filter(protein_Id %in% sample(protein_Id, 100))
+#' istar <-prolfqua::sim_lfq_data_peptide_config()
+#' data <- istar$data
+#'
+#'
 #' lfqdata <- LFQData$new(data, istar$config)
 #' pannot <- ProteinAnnotation$new( lfqdata )
-#'
 #' pannot$annotate_decoys()
 #' pannot$annotate_contaminants()
-#'
-#' stopifnot(pannot$nr_clean(contaminants = FALSE) == 163)
-#' stopifnot(pannot$nr_clean(contaminants = TRUE) == 158)
-#' stopifnot(nrow(pannot$clean(contaminants = FALSE)) == 163)
-#' stopifnot(nrow(pannot$clean(contaminants = TRUE)) == 158)
 #' dd <- pannot$clean()
 #' tmp <- lfqdata$get_subset(dd)
-#' stopifnot(tmp$hierarchy_counts()$protein_Id == 158)
+#'
 #'
 ProteinAnnotation <-
   R6::R6Class("ProteinAnnotation",
@@ -54,6 +49,8 @@ ProteinAnnotation <-
                   } else {
                     self$row_annot <- distinct(select(lfqdata$data, self$pID))
                   }
+
+
                 },
                 #' @description
                 #' annotate rev sequences
diff --git a/R/tidyMS_R6_TransitionCorrelations.R b/R/tidyMS_R6_TransitionCorrelations.R
index 5efc36318..ceb3964dd 100644
--- a/R/tidyMS_R6_TransitionCorrelations.R
+++ b/R/tidyMS_R6_TransitionCorrelations.R
@@ -645,18 +645,14 @@ nr_B_in_A <- function(pdata, config , merge = TRUE){
 #' @keywords internal
 #' @family summary
 #' @examples
-#' bb <- prolfqua::prolfqua_data('data_ionstar')$filtered()
-#' bb$config <- old2new(bb$config)
-#' stopifnot(nrow(bb$data) == 25780)
-#' configur <- bb$config$clone(deep=TRUE)
-#' data <- bb$data
-#'
-#' nr_B_in_A_per_sample(data, configur, nested =FALSE)
-#' bb <- prolfqua::prolfqua_data('data_IonstarProtein_subsetNorm')
-#' bb$config <- old2new(config = bb$config$clone( deep = TRUE))
+#' debug(nr_B_in_A_per_sample)
+#' bb <- prolfqua::sim_lfq_data_peptide_config()
+#' nr_B_in_A_per_sample(bb$data, bb$config, nested =FALSE)
+#' bb <- prolfqua::sim_lfq_data_protein_config()
 #' nr_B_in_A_per_sample(bb$data, bb$config, nested=FALSE)
 #'
 nr_B_in_A_per_sample <- function(data, config, nested = TRUE){
+  #TODO wew check for deprecation since not used.
   cf <- config
 
   levelA <- cf$table$hierarchy_keys_depth()
@@ -687,6 +683,58 @@ nr_B_in_A_per_sample <- function(data, config, nested = TRUE){
 
 
 
+#' Aggregates e.g. protein abundances from peptide abundances
+#'
+#' @export
+#' @examples
+#' dd <- prolfqua::sim_lfq_data_peptide_config()
+#' dd$data <- na.omit(dd$data)
+#' xd <- nr_obs_sample(dd$data, dd$config)
+#' xd
+#' xd$nr_children |> table()
+#'
+#'
+#' dp <- prolfqua::sim_lfq_data_protein_config()
+#' xp <- nr_obs_sample(dp$data, dp$config)
+#' xp$nr_peptides |> table()
+#'
+nr_obs_sample <- function(data, config, new_child = config$table$nr_children){
+  data <- na.omit(data)
+  nr_children <- data |>
+    group_by(!!!rlang::syms(c(config$table$hierarchy_keys_depth(), config$table$fileName))) |>
+    summarize(!!new_child := sum(!!sym(config$table$nr_children), na.rm = TRUE), .groups = "drop")
+  return(nr_children)
+}
+
+#' Aggregates e.g. protein abundances from peptide abundances
+#'
+#' @export
+#' @examples
+#' dd <- prolfqua::sim_lfq_data_peptide_config()
+#'
+#' xd <- nr_obs_hierarchy(dd$data, dd$config)
+#' 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)
+#'
+#'
+nr_obs_hierarchy <- 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()) |>
+      distinct() |>
+      dplyr::group_by(!!sym(tb$hierarchy_keys_depth())) |>
+      dplyr::summarize(!!name_nr_child := dplyr::n(), .groups = "drop")
+  } else {
+    xz <- nr_obs_sample(data,config)
+    xz <- x |> group_by(!!sym(tb$hierarchy_keys_depth())) |>
+      summarize(!!name_nr_child := max(!!sym(tb$nr_children)), .groups = "drop")
+  }
+}
+
+
 # Summarize Intensities by Intensity or NAs ----
 .rankProteinPrecursors <- function(data,
                                    config,
diff --git a/R/tidyMS_aggregation.R b/R/tidyMS_aggregation.R
index 9b45818bd..70c35d2ac 100644
--- a/R/tidyMS_aggregation.R
+++ b/R/tidyMS_aggregation.R
@@ -620,28 +620,14 @@ estimate_intensity <- function(data, config, .func)
     dplyr::ungroup()
 
   new_child = paste0("nr_",config$table$hierarchy_keys_depth())
-  res_nr_children <- nr_obs(data, config, new_child = new_child)
+  res_nr_children <- nr_obs_sample(data, config, new_child = new_child)
   unnested <- inner_join(unnested, res_nr_children, by = c(config$table$hierarchy_keys_depth(), config$table$fileName))
   newconfig$table$nr_children = new_child
   return(list(data = unnested, config = newconfig))
 }
 
-#' Aggregates e.g. protein abundances from peptide abundances
-#'
-#' @export
-#' @examples
-#' dd <- prolfqua::sim_lfq_data_peptide_config()
-#' dd$data <- na.omit(dd$data)
-#' xd <- nr_obs(dd$data, dd$config)
-#'
-#' #xd |> head()
-#'
-#' xd$nr_children |> table()
-nr_obs <- function(data, config, new_child = config$table$nr_children){
-  nr_children <- data |> group_by(!!!rlang::syms(c(config$table$hierarchy_keys_depth(), config$table$fileName))) |>
-    summarize(!!new_child := sum(!!sym(config$table$nr_children), na.rm = TRUE))
-  return(nr_children)
-}
+
+
 
 #' Plot feature data and result of aggregation
 #'
@@ -767,7 +753,7 @@ aggregate_intensity_topN <- function(pdata , config, .func, N = 3){
     hierarchy = config$table$hierarchy[seq_len(config$table$hierarchyDepth)])
 
   new_child_name <- paste0("nr_", config$table$hierarchy_keys_depth() )
-  res_nr_children <- nr_obs(pdata, config, new_child = new_child_name)
+  res_nr_children <- nr_obs_sample(pdata, config, new_child = new_child_name)
   sumTopInt <- inner_join(
     sumTopInt, res_nr_children,
     by = c(config$table$fileName, config$table$hierarchy_keys_depth()))
diff --git a/man/ProteinAnnotation.Rd b/man/ProteinAnnotation.Rd
index 7f8fef2db..d0230aa22 100644
--- a/man/ProteinAnnotation.Rd
+++ b/man/ProteinAnnotation.Rd
@@ -10,22 +10,17 @@ Decorates LFQData with a row annotation and some protein specific functions.
 }
 \examples{
 
-istar <- prolfqua_data('data_ionstar')$filtered()
-istar$config <- old2new(istar$config)
-data <- istar$data #|> dplyr::filter(protein_Id \%in\% sample(protein_Id, 100))
+istar <-prolfqua::sim_lfq_data_peptide_config()
+data <- istar$data
+
+
 lfqdata <- LFQData$new(data, istar$config)
 pannot <- ProteinAnnotation$new( lfqdata )
-
 pannot$annotate_decoys()
 pannot$annotate_contaminants()
-
-stopifnot(pannot$nr_clean(contaminants = FALSE) == 163)
-stopifnot(pannot$nr_clean(contaminants = TRUE) == 158)
-stopifnot(nrow(pannot$clean(contaminants = FALSE)) == 163)
-stopifnot(nrow(pannot$clean(contaminants = TRUE)) == 158)
 dd <- pannot$clean()
 tmp <- lfqdata$get_subset(dd)
-stopifnot(tmp$hierarchy_counts()$protein_Id == 158)
+
 
 }
 \seealso{
diff --git a/man/get_imputed_contrasts.Rd b/man/get_imputed_contrasts.Rd
index 7bce96d97..7a59ab4d5 100644
--- a/man/get_imputed_contrasts.Rd
+++ b/man/get_imputed_contrasts.Rd
@@ -29,13 +29,9 @@ istar <- sim_lfq_data_peptide_config()
 config <- istar$config
 analysis <- istar$data
 data <- complete_cases(analysis, config)
-config$parameter$qVal_individual_threshold <- 0.01
-data <- prolfqua::remove_large_QValues(data, config)
-data <- complete_cases(data, config)
 
 Contrasts <- c("dilution.b-a" = "group_A - group_B", "dilution.c-e" = "group_A - group_Ctrl")
 res <- get_imputed_contrasts(data, config, Contrasts)
-
 config <- config
 contrasts <- Contrasts
 imputed <- missigness_impute_factors_interactions(data, config, value = "imputed" )
diff --git a/man/nr_B_in_A_per_sample.Rd b/man/nr_B_in_A_per_sample.Rd
index e3b402704..0ec952046 100644
--- a/man/nr_B_in_A_per_sample.Rd
+++ b/man/nr_B_in_A_per_sample.Rd
@@ -10,15 +10,10 @@ nr_B_in_A_per_sample(data, config, nested = TRUE)
 how many peptides per protein in each sample
 }
 \examples{
-bb <- prolfqua::prolfqua_data('data_ionstar')$filtered()
-bb$config <- old2new(bb$config)
-stopifnot(nrow(bb$data) == 25780)
-configur <- bb$config$clone(deep=TRUE)
-data <- bb$data
-
-nr_B_in_A_per_sample(data, configur, nested =FALSE)
-bb <- prolfqua::prolfqua_data('data_IonstarProtein_subsetNorm')
-bb$config <- old2new(config = bb$config$clone( deep = TRUE))
+debug(nr_B_in_A_per_sample)
+bb <- prolfqua::sim_lfq_data_peptide_config()
+nr_B_in_A_per_sample(bb$data, bb$config, nested =FALSE)
+bb <- prolfqua::sim_lfq_data_protein_config()
 nr_B_in_A_per_sample(bb$data, bb$config, nested=FALSE)
 
 }
diff --git a/man/nr_obs.Rd b/man/nr_obs.Rd
deleted file mode 100644
index bb23a8248..000000000
--- a/man/nr_obs.Rd
+++ /dev/null
@@ -1,20 +0,0 @@
-% Generated by roxygen2: do not edit by hand
-% Please edit documentation in R/tidyMS_aggregation.R
-\name{nr_obs}
-\alias{nr_obs}
-\title{Aggregates e.g. protein abundances from peptide abundances}
-\usage{
-nr_obs(data, config, new_child = config$table$nr_children)
-}
-\description{
-Aggregates e.g. protein abundances from peptide abundances
-}
-\examples{
-dd <- prolfqua::sim_lfq_data_peptide_config()
-dd$data <- na.omit(dd$data)
-xd <- nr_obs(dd$data, dd$config)
-
-#xd |> head()
-
-xd$nr_children |> table()
-}
diff --git a/man/nr_obs_hierarchy.Rd b/man/nr_obs_hierarchy.Rd
new file mode 100644
index 000000000..78abcfc23
--- /dev/null
+++ b/man/nr_obs_hierarchy.Rd
@@ -0,0 +1,28 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/tidyMS_R6_TransitionCorrelations.R
+\name{nr_obs_hierarchy}
+\alias{nr_obs_hierarchy}
+\title{Aggregates e.g. protein abundances from peptide abundances}
+\usage{
+nr_obs_hierarchy(
+  data,
+  config,
+  from_children = TRUE,
+  name_nr_child = "nr_child_exp"
+)
+}
+\description{
+Aggregates e.g. protein abundances from peptide abundances
+}
+\examples{
+dd <- prolfqua::sim_lfq_data_peptide_config()
+
+xd <- nr_obs_hierarchy(dd$data, dd$config)
+xd
+dp <- prolfqua::sim_lfq_data_protein_config()
+debug(nr_obs_hierarchy)
+nr_obs_sample(dp$data, dp$config)
+xd <- nr_obs_hierarchy(dp$data, dp$config)
+
+
+}
diff --git a/man/nr_obs_sample.Rd b/man/nr_obs_sample.Rd
new file mode 100644
index 000000000..f2af2dc1b
--- /dev/null
+++ b/man/nr_obs_sample.Rd
@@ -0,0 +1,24 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/tidyMS_R6_TransitionCorrelations.R
+\name{nr_obs_sample}
+\alias{nr_obs_sample}
+\title{Aggregates e.g. protein abundances from peptide abundances}
+\usage{
+nr_obs_sample(data, config, new_child = config$table$nr_children)
+}
+\description{
+Aggregates e.g. protein abundances from peptide abundances
+}
+\examples{
+dd <- prolfqua::sim_lfq_data_peptide_config()
+dd$data <- na.omit(dd$data)
+xd <- nr_obs_sample(dd$data, dd$config)
+xd
+xd$nr_children |> table()
+
+
+dp <- prolfqua::sim_lfq_data_protein_config()
+xp <- nr_obs_sample(dp$data, dp$config)
+xp$nr_peptides |> table()
+
+}