Merge pull request #7 from bigbio/dev

major changes about PSMs bugs.

pyproject.toml

@@ -3,7 +3,7 @@ name = "quantms-rescoring"
 description = "quantms-rescoring: Python scripts and helpers for the quantMS workflow"
 readme = "README.md"
 license = "MIT"
-version = "0.0.2"
+version = "0.0.3"
 authors = [
     "Yasset Perez-Riverol <ypriverol@gmail.com>",
     "Dai Chengxin <chengxin2024@126.com>",

quantmsrescore/__init__.py

@@ -1 +1 @@
-__version__ = "0.0.2"
+__version__ = "0.0.3"

quantmsrescore/ms2rescore.py

@@ -36,6 +36,7 @@ def __init__(self, filename: Union[Path, str], *args, **kwargs) -> None:
         self.user_params_metadata = self._get_userparams_metadata(self.peptide_ids[0].getHits()[0])
         self.rescoring_features = self._get_rescoring_features(self.peptide_ids[0].getHits()[0])
         self.skip_invalid_psm = 0
+        self.new_peptide_ids = []
 
     def __iter__(self) -> Iterable[PSM]:
         """
@@ -61,14 +62,24 @@ def __iter__(self) -> Iterable[PSM]:
 
         psm_list: return [PSM 1]
 
+        The invalid PSM are directly from search engines results (MSGF+). The search engine doesn't report search
+        score features (e.g. MSGF:ScoreRatio) for these invalid PSMs. And we can observe the NumMatchedMainIons of
+        peptide hit is 0. So we should remove these invalid PSMs
+
         """
         for peptide_id in self.peptide_ids:
+            new_hits = []
             for peptide_hit in peptide_id.getHits():
                 psm = self._parse_psm(self.protein_ids, peptide_id, peptide_hit)
                 if psm is not None:
+                    new_hits.append(peptide_hit)
                     yield psm
                 else:
                     self.skip_invalid_psm += 1
+            # If it is a valid Peptide Hits then keep it
+            if len(new_hits) > 0:
+                peptide_id.setHits(new_hits)
+                self.new_peptide_ids.append(peptide_id)
 
     def _parse_psm(
             self,
@@ -239,12 +250,15 @@ def rescore_idxml(input_file, output_file, config) -> None:
         logging.warning(
             f"Removed {reader.skip_invalid_psm} PSMs without search engine features!"
         )
-
+        # Synchronised acquisition of new peptide IDs after removing invalid PSMs
+        peptide_ids = reader.new_peptide_ids
+    else:
+        peptide_ids = reader.peptide_ids
     # Rescore
     rescore(config, psm_list)
 
     # Filter out PeptideHits within PeptideIdentification(s) that could not be processed by all feature generators
-    peptide_ids_filtered = filter_out_artifact_psms(psm_list, reader.peptide_ids)
+    peptide_ids_filtered = filter_out_artifact_psms(psm_list, peptide_ids)
 
     # Write
     writer = IdXMLWriter(output_file, reader.protein_ids, peptide_ids_filtered)

recipe/meta.yaml

@@ -1,7 +1,7 @@
 # recipe/meta.yaml
 package:
   name: quantms-rescoring
-  version: "0.0.2"
+  version: "0.0.3"
 
 source:
   path: ../