Remove redundant GENCODE IDs

GENCODE IDs are currently only unique when combining the protein (ENSP) and transcript (ENST) IDs. Since there are so few duplicates, we will remove them rather than concatenating these IDs in the "protein_id" column of the output of make_results_ratio_* and make_rii_peptide_* functions.
PNNL-Comp-Mass-Spec · Jan 10, 2023 · 203a8ef · 203a8ef
1 parent cc927db
commit 203a8ef
Show file tree

Hide file tree

Showing 3 changed files with 41 additions and 73 deletions.
diff --git a/NAMESPACE b/NAMESPACE
@@ -72,13 +72,15 @@ importFrom(dplyr,inner_join)
 importFrom(dplyr,left_join)
 importFrom(dplyr,mutate)
 importFrom(dplyr,n)
+importFrom(dplyr,pull)
 importFrom(dplyr,rename)
 importFrom(dplyr,select)
 importFrom(dplyr,starts_with)
 importFrom(dplyr,summarise)
 importFrom(dplyr,summarize)
 importFrom(dplyr,ungroup)
 importFrom(plyr,llply)
+importFrom(purrr,map)
 importFrom(purrr,reduce)
 importFrom(readr,read_tsv)
 importFrom(tibble,rownames_to_column)

diff --git a/R/motrpac_bic_funtions.R b/R/motrpac_bic_funtions.R
@@ -133,11 +133,6 @@ make_rii_peptide_gl <- function(msnid,
   aggregation_level <- c("accession", "peptide")
   annotation <- toupper(annotation)
 
-  if(annotation == "GENCODE"){
-    psms(msnid) <- psms(msnid) %>%
-      mutate(accession = sub("(ENSP[^\\|]+\\|ENST[^\\|]+).*",
-                             "\\1", accession))
-  }
   crosstab <- create_crosstab(msnid,
                               masic_data,
                               aggregation_level,
@@ -146,7 +141,6 @@ make_rii_peptide_gl <- function(msnid,
     as.data.frame() %>%
     rownames_to_column("Specie")
 
-
   ## Fetch conversion table
   from <- annotation
   to <- c("SYMBOL", "ENTREZID")
@@ -163,7 +157,7 @@ make_rii_peptide_gl <- function(msnid,
     from <- "SYMBOL"
     to <- "ENTREZID"
   } else if (annotation == "GENCODE") {
-    rgx <- "(ENSP[^\\|]+\\|ENST[^\\|]+).*"
+    rgx <- "(ENSP[^\\|]+).*"
     grp <- "\\1"
     fasta_names <- parse_FASTA_names(fasta_file, "gencode") %>%
       dplyr::rename(SYMBOL = gene)
@@ -175,14 +169,6 @@ make_rii_peptide_gl <- function(msnid,
     fetch_conversion_table(org_name, from = from, to = to)
   )
 
-  # Add ENTREZID column to parse_FASTA_names results
-  if (annotation == "GENCODE") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL") %>%
-      mutate(protein_id = paste(protein_id, transcript_id, sep = "|"))
-  } else if (annotation == "UNIPROT") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL")
-  }
-
   # Feature data
   feature_data <- crosstab %>%
     dplyr::select(Specie) %>%
@@ -196,7 +182,9 @@ make_rii_peptide_gl <- function(msnid,
       left_join(conv, by = c("ANNOTATION" = annotation)) %>%
       dplyr::select(-ANNOTATION)
   } else if (annotation %in% c("GENCODE", "UNIPROT")) {
-    feature_data <- left_join(feature_data, tab, by = "protein_id")
+    feature_data <- left_join(feature_data, fasta_names,
+                              by = "protein_id") %>%
+      left_join(conv, by = "SYMBOL")
   }
 
   feature_data <- dplyr::rename(feature_data,
@@ -232,8 +220,6 @@ utils::globalVariables(
 )
 
 
-
-
 #' @export
 #' @rdname motrpac_bic_output
 make_results_ratio_gl <- function(msnid,
@@ -250,10 +236,6 @@ make_results_ratio_gl <- function(msnid,
   aggregation_level <- c("accession")
   annotation <- toupper(annotation)
 
-  if (annotation == "GENCODE") {
-    msnid$accession <- sub("(ENSP[^\\|]+\\|ENST[^\\|]+).*", "\\1", msnid$accession)
-  }
-
   crosstab <- create_crosstab(msnid, masic_data,
                               aggregation_level,
                               fractions, samples, references) %>%
@@ -275,7 +257,7 @@ make_results_ratio_gl <- function(msnid,
     from <- "SYMBOL"
     to <- "ENTREZID"
   } else if (annotation == "GENCODE") {
-    rgx <- "(ENSP[^\\|]+\\|ENST[^\\|]+).*"
+    rgx <- "(ENSP[^\\|]+).*"
     grp <- "\\1"
     fasta_names <- parse_FASTA_names(fasta_file, "gencode") %>%
       dplyr::rename(SYMBOL = gene)
@@ -287,13 +269,6 @@ make_results_ratio_gl <- function(msnid,
     fetch_conversion_table(org_name, from = from, to = to)
   )
 
-  if (annotation == "GENCODE") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL") %>%
-      mutate(protein_id = paste(protein_id, transcript_id, sep = "|"))
-  } else if (annotation == "UNIPROT") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL")
-  }
-
   # Create Feature data
   feature_data <- crosstab %>%
     dplyr::select(protein_id) %>%
@@ -304,15 +279,16 @@ make_results_ratio_gl <- function(msnid,
       mutate(ANNOTATION = sub(rgx, grp, protein_id)) %>%
       left_join(conv, by = c("ANNOTATION" = annotation)) %>%
       select(-ANNOTATION)
-  } else if (annotation %in% c("UNIPROT", "GENCODE")) {
-    feature_data <- left_join(feature_data, tab, by = "protein_id")
+  } else if (annotation %in% c("GENCODE", "UNIPROT")) {
+    feature_data <- left_join(feature_data, fasta_names,
+                              by = "protein_id") %>%
+      left_join(conv, by = "SYMBOL")
   }
 
   feature_data <- dplyr::rename(feature_data,
                                 gene_symbol = SYMBOL,
                                 entrez_id = ENTREZID)
 
-
   ## Additional info from MS/MS -------------------------------------------
   ids <- psms(msnid) %>%
     dplyr::select(accession, peptide,
@@ -340,7 +316,6 @@ utils::globalVariables(c("noninferableProteins", "percentAACoverage",
                          "percent_coverage", "feature", "transcript_id"))
 
 
-
 #' @export
 #' @rdname motrpac_bic_output
 make_rii_peptide_ph <- function(msnid,
@@ -368,9 +343,6 @@ make_rii_peptide_ph <- function(msnid,
   ## Create Crosstab
   annotation <- toupper(annotation)
 
-  if (annotation == "GENCODE") {
-    msnid$accession = sub("(ENSP[^\\|]+\\|ENST[^\\|]+).*", "\\1", msnid$accession)
-  }
   aggregation_level <- c("accession", "peptide", "SiteID")
   crosstab <- create_crosstab(msnid,
                               masic_data,
@@ -396,7 +368,7 @@ make_rii_peptide_ph <- function(msnid,
     from <- "SYMBOL"
     to <- "ENTREZID"
   } else if (annotation == "GENCODE") {
-    rgx <- "(ENSP[^\\|]+\\|ENST[^\\|]+).*"
+    rgx <- "(ENSP[^\\|]+).*"
     grp <- "\\1"
     fasta_names <- parse_FASTA_names(fasta_file, "gencode") %>%
       dplyr::rename(SYMBOL = gene)
@@ -408,13 +380,6 @@ make_rii_peptide_ph <- function(msnid,
     fetch_conversion_table(org_name, from = from, to = to)
   )
 
-  if (annotation == "GENCODE"){
-    tab <- left_join(fasta_names, conv, by = "SYMBOL") %>%
-      mutate(protein_id = paste(protein_id, transcript_id, sep = "|"))
-  } else if (annotation == "UNIPROT") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL")
-  }
-
   ## Create RII peptide table
   feature_data <- crosstab %>%
     dplyr::select(Specie) %>%
@@ -429,7 +394,9 @@ make_rii_peptide_ph <- function(msnid,
       left_join(conv, by = c("ANNOTATION" = annotation)) %>%
       dplyr::select(-ANNOTATION)
   } else if (annotation %in% c("GENCODE", "UNIPROT")) {
-    feature_data <- left_join(feature_data, tab, by = "protein_id")
+    feature_data <- left_join(feature_data, fasta_names,
+                              by = "protein_id") %>%
+      left_join(conv, by = "SYMBOL")
   }
 
   feature_data <- dplyr::rename(feature_data,
@@ -474,8 +441,6 @@ utils::globalVariables(
 )
 
 
-
-
 #' @export
 #' @rdname motrpac_bic_output
 make_results_ratio_ph <- function(msnid,
@@ -487,8 +452,9 @@ make_results_ratio_ph <- function(msnid,
                                   fasta_file)
 {
   aggregation_level <- c("accession", "SiteID")
-  crosstab <- create_crosstab(msnid, masic_data, aggregation_level, fractions,
-                              samples, references)
+  crosstab <- create_crosstab(msnid, masic_data,
+                              aggregation_level,
+                              fractions, samples, references)
   crosstab <- as.data.frame(crosstab) %>%
     rownames_to_column("Specie")
 
@@ -509,7 +475,7 @@ make_results_ratio_ph <- function(msnid,
     from <- "SYMBOL"
     to <- "ENTREZID"
   } else if (annotation == "GENCODE") {
-    rgx <- "(ENSP[^\\|]+\\|ENST[^\\|]+).*"
+    rgx <- "(ENSP[^\\|]+).*"
     grp <- "\\1"
     fasta_names <- parse_FASTA_names(fasta_file, "gencode") %>%
       dplyr::rename(SYMBOL = gene)
@@ -521,13 +487,6 @@ make_results_ratio_ph <- function(msnid,
     fetch_conversion_table(org_name, from = from, to = to)
   )
 
-  if (annotation == "GENCODE") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL") %>%
-      mutate(protein_id = paste(protein_id, transcript_id, sep = "|"))
-  } else if (annotation == "UNIPROT") {
-    tab <- left_join(fasta_names, conv, by = "SYMBOL")
-  }
-
   ## Create RII peptide table
   feature_data <- crosstab %>%
     select(Specie) %>%
@@ -541,7 +500,9 @@ make_results_ratio_ph <- function(msnid,
       left_join(conv, by = c("ANNOTATION" = annotation)) %>%
       select(-ANNOTATION)
   } else if (annotation %in% c("GENCODE", "UNIPROT")) {
-    feature_data <- left_join(feature_data, tab, by = "protein_id")
+    feature_data <- left_join(feature_data, fasta_names,
+                              by = "protein_id") %>%
+      left_join(conv, by = "SYMBOL")
   }
 
   feature_data <- dplyr::rename(feature_data,
@@ -585,8 +546,6 @@ utils::globalVariables(
 )
 
 
-
-
 #' @export
 #' @rdname motrpac_bic_output
 assess_redundant_protein_matches <- function(msnid, collapse="|") {
@@ -602,8 +561,6 @@ assess_redundant_protein_matches <- function(msnid, collapse="|") {
 }
 
 
-
-
 #' @export
 #' @rdname motrpac_bic_output
 assess_noninferable_proteins <- function(msnid, collapse="|") {

diff --git a/R/run_plexedpiper.R b/R/run_plexedpiper.R
@@ -52,10 +52,10 @@
 #' @importFrom MSnID psms MSnID compute_accession_coverage
 #'   correct_peak_selection extract_sequence_window
 #'   infer_parsimonious_accessions map_mod_sites
-#' @importFrom dplyr %>% full_join select
+#' @importFrom dplyr %>% full_join select mutate filter pull
 #' @importFrom tidyselect where
 #' @importFrom data.table rbindlist
-#' @importFrom purrr reduce
+#' @importFrom purrr reduce map
 #'
 #' @examples \dontrun{
 #' # Example with pseudo-paths
@@ -126,22 +126,20 @@ run_plexedpiper <- function(msgf_output_folder,
 
   suppressMessages(msnid <- MSnID())
 
-  psms(msnid) <- lapply(msgf_output_folder, function(msgf_folder_i) {
+  psms(msnid) <- map(msgf_output_folder, function(msgf_folder_i) {
     out <- read_msgf_data(msgf_folder_i)
     out <- psms(out)
     return(out)
   }) %>%
     rbindlist(fill = TRUE)
 
   if (!is.null(ascore_output_folder)) {
-    ascore <- lapply(ascore_output_folder, function(ascore_folder_i) {
-      read_AScore_results(ascore_folder_i)
-    }) %>%
+    ascore <- map(ascore_output_folder, read_AScore_results) %>%
       rbindlist(fill = TRUE)
   }
 
   if (verbose) {message("- Filtering MASIC results.")}
-  masic_data <- lapply(masic_output_folder, function(masic_folder) {
+  masic_data <- map(masic_output_folder, function(masic_folder) {
     read_masic_data(masic_folder, interference_score = TRUE) %>%
       filter_masic_data(0.5, 0)
   })
@@ -184,8 +182,19 @@ run_plexedpiper <- function(msgf_output_folder,
   msnid <- apply_filter(msnid, "!isDecoy")
 
   if (annotation == "GENCODE") {
-    msnid$accession <- sub("(ENSP[^\\|]+\\|ENST[^\\|]+).*", "\\1", msnid$accession)
-    names(fst) <- sub("(ENSP[^\\|]+\\|ENST[^\\|]+).*", "\\1", names(fst))
+    # Remove duplicate GENCODE protein IDs using FASTA file headers
+    pttrn <- "(ENSP[^\\|]+).*"
+    unique_ids <- data.frame(id_orig = names(fst)) %>%
+      mutate(id_new = sub(pttrn, "\\1", id_orig),
+             duped = duplicated(id_new)) %>%
+      filter(!duped) %>%
+      pull(id_orig)
+    msnid <- apply_filter(msnid, "accession %in% unique_ids")
+    msnid$accession <- sub(pttrn, "\\1", msnid$accession)
+    fst <- fst[names(fst) %in% unique_ids]
+    names(fst) <- sub(pttrn, "\\1", names(fst))
+
+    # Sanity check
     if (anyDuplicated(names(fst)) != 0) {
       stop("Duplicate FASTA entry names!")
     }
@@ -228,7 +237,7 @@ run_plexedpiper <- function(msgf_output_folder,
       stop("Proteomics variable not supported.")
     }
 
-    msnid <- map_mod_sites(msnid,
+    msnid <- map_mod_sites(object          = msnid,
                            fasta           = fst,
                            accession_col   = "accession",
                            peptide_mod_col = "peptide",