Merge pull request #140 from CompOmics/fix/various-io-fixes

Various fixes in io modules

Author: RalfG

psm_utils/__init__.py

@@ -1,6 +1,6 @@
 """Common utilities for parsing and handling PSMs, and search engine results."""
 
-__version__ = "1.5.0.post1"
+__version__ = "1.5.1"
 __all__ = ["Peptidoform", "PSM", "PSMList"]
 
 from warnings import filterwarnings

psm_utils/io/__init__.py

@@ -241,8 +241,16 @@ def _supports_write_psm(writer: type[WriterBase]) -> bool:
         temp_file.close()
         Path(temp_file.name).unlink()
         example_psm = PSM(peptidoform="ACDE", spectrum_id="0")
+
+        # Prepare writer-specific kwargs for writers that need them
+        writer_kwargs = {}
+        if writer == percolator.PercolatorTabWriter:
+            writer_kwargs["style"] = "pin"
+
         try:
-            with writer(temp_file.name, example_psm=example_psm) as writer_instance:
+            with writer(
+                temp_file.name, example_psm=example_psm, **writer_kwargs
+            ) as writer_instance:
                 writer_instance.write_psm(example_psm)
         except NotImplementedError:
             supports_write_psm = False

psm_utils/io/idxml.py

@@ -34,8 +34,11 @@
     import pyopenms as oms  # type: ignore[import]
 
     _has_openms = True
+    # Check if we have pyOpenMS 3.5+ with PeptideIdentificationList
+    _has_peptide_id_list = hasattr(oms, "PeptideIdentificationList")
 except ImportError:
     _has_openms = False
+    _has_peptide_id_list = False
     oms = None  # type: ignore[assignment]
 
 logger = logging.getLogger(__name__)
@@ -157,8 +160,17 @@ def _parse_idxml(self) -> tuple[Any, Any]:
 
         """
         protein_ids: Any = []  # list[oms.ProteinIdentification]
-        peptide_ids: Any = []  # list[oms.PeptideIdentification]
-        oms.IdXMLFile().load(str(self.filename), protein_ids, peptide_ids)  # type: ignore
+        # In pyOpenMS 3.5+, peptide_ids must be a PeptideIdentificationList
+        if _has_peptide_id_list:
+            peptide_ids: Any = oms.PeptideIdentificationList()  # type: ignore
+        else:
+            peptide_ids = []  # list[oms.PeptideIdentification] for pyOpenMS <3.5
+
+        # Load the idXML file - the lists will be populated by pyOpenMS
+        idxml_file = oms.IdXMLFile()  # type: ignore
+        # Ensure filename is a string, not a Path object
+        filename_str: str = str(self.filename)
+        idxml_file.load(filename_str, protein_ids, peptide_ids)
 
         if len(protein_ids) == 0:
             raise IdXMLReaderEmptyListException(
@@ -564,7 +576,10 @@ def _update_existing_ids(
 
             peptide_id.setHits(updated_peptide_hits)
 
-        oms.IdXMLFile().store(str(self.filename), self.protein_ids, self.peptide_ids)  # type: ignore
+        # Store the idXML file
+        idxml_file = oms.IdXMLFile()  # type: ignore
+        filename_str: str = str(self.filename)
+        idxml_file.store(filename_str, self.protein_ids, self.peptide_ids)
 
     def _update_peptide_hit(self, peptide_hit: Any, psm: PSM) -> None:
         """Inplace update of PeptideHit with novel predicted features information from PSM."""
@@ -594,7 +609,11 @@ def _create_ids_for_collection(
     ) -> None:
         """Create ProteinIdentification and PeptideIdentification objects for a single collection."""
         self.protein_ids = [oms.ProteinIdentification()]  # type: ignore
-        self.peptide_ids = []
+        # In pyOpenMS 3.5+, peptide_ids must be a PeptideIdentificationList
+        if _has_peptide_id_list:
+            self.peptide_ids = oms.PeptideIdentificationList()  # type: ignore
+        else:
+            self.peptide_ids = []  # list[oms.PeptideIdentification] for pyOpenMS <3.5
 
         # Set msrun filename with spectra_data meta value
         msrun_reference = [str(run).encode() for run in runs.keys()]
@@ -617,14 +636,19 @@ def _create_ids_for_collection(
                 # Create PeptideHits
                 peptide_hits = [self._create_peptide_hit(psm) for psm in psms]
                 peptide_id.setHits(peptide_hits)
-                self.peptide_ids.append(peptide_id)
+                # Use push_back for pyOpenMS 3.5+, append for older versions
+                if _has_peptide_id_list:
+                    self.peptide_ids.push_back(peptide_id)  # type: ignore
+                else:
+                    self.peptide_ids.append(peptide_id)  # type: ignore[union-attr]
 
         # Create protein hits
         self._create_protein_hits(protein_list)
 
         # Write idXML file
-        filename = "/".join(filter(None, [collection, str(self.filename)]))
-        oms.IdXMLFile().store(filename, self.protein_ids, self.peptide_ids)  # type: ignore
+        filename: str = "/".join(filter(None, [collection, str(self.filename)]))
+        idxml_file = oms.IdXMLFile()  # type: ignore
+        idxml_file.store(filename, self.protein_ids, self.peptide_ids)  # type: ignore
 
     def _create_peptide_identification(
         self,

psm_utils/io/peptide_record.py

@@ -448,7 +448,8 @@ def from_dataframe(peprec_df: pd.DataFrame) -> PSMList:
     """
     psm_list = []
     for _, row in peprec_df.iterrows():
-        entry = _PeprecEntry(**row.to_dict())
+        row_dict = {str(k): v for k, v in row.to_dict().items()}
+        entry = _PeprecEntry(**row_dict)
         psm_list.append(PeptideRecordReader._entry_to_psm(entry, filename=""))
     return PSMList(psm_list=psm_list)
 

psm_utils/io/pepxml.py

@@ -27,6 +27,12 @@
     "mzFidelity",
 ]
 
+KNOWN_METADATA_KEYS = [
+    "num_matched_ions",
+    "tot_num_ions",
+    "num_missed_cleavages",
+]
+
 
 class PepXMLReader(ReaderBase):
     """Reader for pepXML PSM files."""
@@ -127,47 +133,51 @@ def _parse_peptidoform(
 
     def _parse_psm(self, spectrum_query: dict[str, Any], search_hit: dict[str, Any]) -> PSM:
         """Parse pepXML PSM to PSM."""
-        metadata = {
-            "num_matched_ions": str(search_hit["num_matched_ions"]),
-            "tot_num_ions": str(search_hit["tot_num_ions"]),
-            "num_missed_cleavages": str(search_hit["num_missed_cleavages"]),
-        }
+        # Build metadata from optional search hit fields
+        metadata = {key: str(search_hit[key]) for key in KNOWN_METADATA_KEYS if key in search_hit}
+
+        # Add all search scores to metadata
         metadata.update(
             {
-                f"search_score_{key.lower()}": str(search_hit["search_score"][key])
-                for key in search_hit["search_score"]
+                f"search_score_{key.lower()}": str(value)
+                for key, value in search_hit["search_score"].items()
             }
         )
 
+        # Build provenance data from optional spectrum query fields
+        provenance_data = {
+            k: str(v)
+            for k, v in {
+                "pepxml_index": spectrum_query.get("index"),
+                "start_scan": spectrum_query.get("start_scan"),
+                "end_scan": spectrum_query.get("end_scan"),
+            }.items()
+            if v is not None
+        }
+
         return PSM(
             peptidoform=self._parse_peptidoform(
                 search_hit["peptide"],
                 search_hit["modifications"],
                 spectrum_query["assumed_charge"],
             ),
-            spectrum_id=spectrum_query["spectrumNativeID"]
-            if "spectrumNativeID" in spectrum_query
-            else spectrum_query["spectrum"],
+            spectrum_id=spectrum_query.get("spectrumNativeID", spectrum_query.get("spectrum")),
             run=None,
             collection=None,
             spectrum=None,
             is_decoy=None,
-            score=search_hit["search_score"][self.score_key],
+            score=search_hit["search_score"].get(self.score_key, None),
             qvalue=None,
             pep=None,
             precursor_mz=mass_to_mz(
                 spectrum_query["precursor_neutral_mass"], spectrum_query["assumed_charge"]
             ),
             retention_time=spectrum_query.get("retention_time_sec"),
             ion_mobility=spectrum_query.get("ion_mobility"),
-            protein_list=[p["protein"] for p in search_hit["proteins"]],
-            rank=search_hit["hit_rank"],
+            protein_list=[p["protein"] for p in search_hit.get("proteins", [])],
+            rank=search_hit.get("hit_rank", None),
             source=None,
-            provenance_data={
-                "pepxml_index": str(spectrum_query["index"]),
-                "start_scan": str(spectrum_query["start_scan"]),
-                "end_scan": str(spectrum_query["end_scan"]),
-            },
+            provenance_data=provenance_data,
             metadata=metadata,
             rescoring_features={},
         )

psm_utils/io/percolator.py

@@ -342,8 +342,10 @@ def write_file(self, psm_list: PSMList) -> None:
                 f, fieldnames=self._columns, delimiter="\t", extrasaction="ignore"
             )
             writer.writeheader()
-            for psm in psm_list:
-                writer.writerow(self._psm_to_entry(psm))
+            for i, psm in enumerate(psm_list):
+                entry = self._psm_to_entry(psm)
+                entry["ScanNr"] = i
+                writer.writerow(entry)
 
     def _psm_to_entry(self, psm: PSM) -> dict[str, Any]:
         """Parse PSM to Percolator Tab entry."""