speeds up add_peptide_positions by dropping Biostrings::match (#53)

CambridgeCentreForProteomics · Jul 1, 2024 · 197a446 · 197a446
1 parent 2b92ec4
commit 197a446
Showing 1 changed file with 4 additions and 3 deletions.
diff --git a/R/ptm.R b/R/ptm.R
@@ -323,8 +323,10 @@ add_peptide_positions <- function(obj,
     if (!protein %in% names(proteome)) {
       return(c(NA, NA))
     }
-    peptide_start <- Biostrings::start(Biostrings::matchPattern(sequence, proteome[[protein]]))
-    peptide_end <- Biostrings::end(Biostrings::matchPattern(sequence, proteome[[protein]]))
+
+    m <- gregexpr(sprintf('(?=(%s))', sequence), proteome[[protein]], perl=T)[[1]]
+    peptide_start <- attr(m,"capture.start")
+    peptide_end <- attr(m,"capture.start") + attr(m,"capture.length") - 1
 
     if (length(peptide_start) != 1) {
       return(c(NA, NA))
@@ -335,7 +337,6 @@ add_peptide_positions <- function(obj,
     }
   }
 
-
   obj[, c('peptide_start', 'peptide_end')] <- t(apply(
     obj,
     MARGIN = 1, function(x) combine_protein_peptide_positions(