Fix Fragpipe proteoform parsing

psm_utils/io/fragpipe.py

@@ -17,6 +17,7 @@
 from abc import ABC
 from pathlib import Path
 from typing import Iterable, Optional
+from pyteomics.proforma import MassModification, to_proforma
 
 from psm_utils.io._base_classes import ReaderBase
 from psm_utils.io._utils import set_csv_field_size_limit
@@ -76,7 +77,7 @@ def _get_peptide_spectrum_match(self, psm_dict) -> PSM:
 
         return PSM(
             peptidoform=self._parse_peptidoform(
-                psm_dict["Modified Peptide"], psm_dict["Peptide"], psm_dict["Charge"]
+                psm_dict["Peptide"], psm_dict["Assigned Modifications"], psm_dict["Charge"]
             ),
             spectrum_id=self._parse_spectrum_id(psm_dict["Spectrum"]),
             run=self._parse_run(psm_dict["Spectrum File"]),
@@ -98,22 +99,24 @@ def _get_peptide_spectrum_match(self, psm_dict) -> PSM:
         )
 
     @staticmethod
-    def _parse_peptidoform(mod_peptide: str, peptide: str, charge: Optional[str]) -> str:
+    def _parse_peptidoform(peptide: str, modifications: str, charge: Optional[str]) -> str:
         """Parse the peptidoform from the modified peptide, peptide, and charge columns."""
-        if mod_peptide:
-            peptide = mod_peptide
-            # N-terminal modification
-            if peptide.startswith("n"):
-                peptide = peptide[1:]
-                # A hyphen needs to be added after the N-terminal modification, thus after the ]
-                peptide = peptide.replace("]", "]-", 1)
-            # C-terminal modification
-            if peptide.endswith("]"):
-                if "c[" in peptide:
-                    peptide = peptide.replace("c[", "-[", 1)
-        if charge:
-            peptide += f"/{int(float(charge))}"
-        return peptide
+        sequence = [(aa, []) for aa in peptide]
+        n_term, c_term = [], []
+        for mod_entry in modifications.split(", "):
+            if mod_entry:
+                site, mass = mod_entry[:-1].split("(")
+                mass = float(mass)
+                if site == "N-term":
+                    n_term.append(MassModification(mass))
+                elif site == "C-term":
+                    c_term.append(MassModification(mass))
+                else:
+                    res = site[-1]
+                    idx = int(site[:-1]) - 1
+                    assert sequence[idx][0] == res
+                    sequence[idx][1].append(MassModification(mass))
+        return to_proforma(sequence, n_term=n_term, c_term=c_term, charge_state=charge)
 
     @staticmethod
     def _parse_spectrum_id(spectrum: str) -> str:

tests/test_io/test_fragpipe.py

@@ -42,16 +42,14 @@ def test_iter(self):
 
     def test__parse_peptidoform(self):
         test_cases = [
-            (("LHM[147]TNQNMEKc[17]", "LHMTNQNMEK", "3"), "LHM[147]TNQNMEK-[17]/3"),
-            (("n[43]ANIAVQR", "ANIAVQR", "2"), "[43]-ANIAVQR/2"),
-            ((None, "IPAVTYPK", "2"), "IPAVTYPK/2"),
-            (("", "IPAVTYPK", "2"), "IPAVTYPK/2"),
-            (("", "IPAVTYPK", 2), "IPAVTYPK/2"),
+            (("LHMTNQNMEK", "3M(15.994915), C-term(17.034480)", "3"), "LHM[+15.9949]TNQNMEK-[+17.0345]/3"),
+            (("ANIAVQR", "N-term(42.0106)", "2"), "[+42.0106]-ANIAVQR/2"),
+            (("IPAVTYPK", "", "2"), "IPAVTYPK/2"),
         ]
 
         reader = FragPipeReader("./tests/test_data/test_fragpipe.tsv")
-        for (peptide, modified_peptide, charge), expected in test_cases:
-            assert reader._parse_peptidoform(peptide, modified_peptide, charge) == expected
+        for (peptide, modifications, charge), expected in test_cases:
+            assert reader._parse_peptidoform(peptide, modifications, charge) == expected
 
     def test__parse_spectrum_id(self):
         test_cases = [