introduce ModifiedSequenceReader

alphabase/constants/const_files/psm_reader.yaml

@@ -191,6 +191,8 @@ spectronaut_report:
     'charge': 'charge'
   mod_seq_columns:
     - 'ModifiedSequence'
+  precursor_id_columns:
+    - "EG.PrecursorId"
   modification_mapping_type: 'maxquant'
 
 spectronaut:
@@ -215,6 +217,8 @@ spectronaut:
     - 'ModifiedPeptideSequence'
     - 'LabeledSequence'
     - 'FullUniModPeptideName'
+  precursor_id_columns:
+    - "EG.PrecursorId"
   modification_mapping_type: 'maxquant'
 
 library_reader_base:

alphabase/psm_reader/__init__.py

@@ -11,6 +11,7 @@
     "MSFragger_PSM_TSV_Reader",
     "pFindReader",
     "MSFraggerPepXML",
+    "MSFraggerPepXMLReader",
     "SageReaderTSV",
     "SageReaderParquet",
 ]
@@ -39,6 +40,7 @@
 from alphabase.psm_reader.msfragger_reader import (
     MSFragger_PSM_TSV_Reader,
     MSFraggerPepXML,
+    MSFraggerPepXMLReader,
 )
 from alphabase.psm_reader.msfragger_reader import (
     register_readers as register_fragger_readers,

alphabase/psm_reader/alphapept_reader.py

@@ -76,21 +76,15 @@ def _pre_process(self, df: pd.DataFrame) -> pd.DataFrame:
         df[PsmDfCols.CHARGE] = df[PsmDfCols.CHARGE].astype(int)
         return df
 
-    def _load_modifications(self, df: pd.DataFrame) -> None:
-        if len(df) == 0:
-            self._psm_df[PsmDfCols.SEQUENCE] = ""
-            self._psm_df[PsmDfCols.MODS] = ""
-            self._psm_df[PsmDfCols.MOD_SITES] = ""
-            self._psm_df[PsmDfCols.DECOY] = 0
-            return
-
+    def _load_modifications(self, origin_df: pd.DataFrame) -> None:
         (
             self._psm_df[PsmDfCols.SEQUENCE],
             self._psm_df[PsmDfCols.MODS],
             self._psm_df[PsmDfCols.MOD_SITES],
             _charges,
             self._psm_df[PsmDfCols.DECOY],
-        ) = zip(*df["precursor"].apply(parse_ap))
+        ) = zip(*origin_df["precursor"].apply(parse_ap))
+
         self._psm_df[PsmDfCols.DECOY] = self._psm_df[PsmDfCols.DECOY].astype(np.int8)
 
 

alphabase/psm_reader/dia_psm_reader.py

@@ -4,16 +4,12 @@
 import pandas as pd
 
 from alphabase.psm_reader.keys import PsmDfCols
-from alphabase.psm_reader.maxquant_reader import MaxQuantReader
+from alphabase.psm_reader.maxquant_reader import ModifiedSequenceReader
 from alphabase.psm_reader.psm_reader import psm_reader_provider
 
 
-class SpectronautReader(MaxQuantReader):
-    """Reader for Spectronaut's output library TSV/CSV.
-
-    Other parameters, please see `MaxQuantReader`
-    in `alphabase.psm_reader.maxquant_reader`
-    """
+class SpectronautReader(ModifiedSequenceReader):
+    """Reader for Spectronaut's output library TSV/CSV."""
 
     _reader_type = "spectronaut"
     _add_unimod_to_mod_mapping = True
@@ -39,34 +35,23 @@ class SwathReader(SpectronautReader):
     _add_unimod_to_mod_mapping = True
 
 
-class DiannReader(MaxQuantReader):
+class DiannReader(ModifiedSequenceReader):
     """Reader for DIANN data."""
 
     _reader_type = "diann"
     _add_unimod_to_mod_mapping = True
     _min_max_rt_norm = False
 
-    def _pre_process(self, df: pd.DataFrame) -> pd.DataFrame:
-        """DIANN-specific preprocessing of output data.
-
-        Nothing to do for DIANN, still method of superclass needs to be overwritten.
-        TODO disentangle the inheritance structure.
-        """
-        return df
-
-    def _post_process(self) -> None:
-        super()._post_process()
+    def _post_process(self, origin_df: pd.DataFrame) -> None:
         self._psm_df.rename(
             columns={PsmDfCols.SPEC_IDX: PsmDfCols.DIANN_SPEC_INDEX}, inplace=True
         )
 
+        super()._post_process(origin_df)
 
-class SpectronautReportReader(MaxQuantReader):
-    """Reader for Spectronaut's report TSV/CSV.
 
-    Other parameters, please see `MaxQuantReader`
-    in `alphabase.psm_reader.maxquant_reader`
-    """
+class SpectronautReportReader(ModifiedSequenceReader):
+    """Reader for Spectronaut's report TSV/CSV."""
 
     _reader_type = "spectronaut_report"
     _add_unimod_to_mod_mapping = True
@@ -75,7 +60,7 @@ class SpectronautReportReader(MaxQuantReader):
     def _pre_process(self, df: pd.DataFrame) -> pd.DataFrame:
         """Spectronaut report-specific preprocessing of output data."""
         df[[self.mod_seq_column, PsmDfCols.CHARGE]] = df[
-            "EG.PrecursorId"  # TODO: move to yaml
+            self._precursor_id_column
         ].str.split(".", expand=True, n=2)
         df[PsmDfCols.CHARGE] = df[PsmDfCols.CHARGE].astype(np.int8)
         return df

alphabase/psm_reader/maxquant_reader.py

@@ -1,6 +1,7 @@
 """Reader for MaxQuant data."""
 
 import warnings
+from abc import ABC
 from typing import List, Optional
 
 import numba
@@ -123,10 +124,9 @@ def parse_mod_seq(
     )
 
 
-class MaxQuantReader(PSMReaderBase):
-    """Reader for MaxQuant data."""
+class ModifiedSequenceReader(PSMReaderBase, ABC):
+    """Reader for MaxQuant-like data."""
 
-    _reader_type = "maxquant"
     _add_unimod_to_mod_mapping = True
 
     def __init__(  # noqa: PLR0913, D417 # too many arguments in function definition, missing argument descriptions
@@ -142,7 +142,7 @@ def __init__(  # noqa: PLR0913, D417 # too many arguments in function definition
         fixed_C57: Optional[bool] = None,  # noqa: N803 TODO: make this  *,fixed_c57  (breaking)
         **kwargs,
     ):
-        """Reader for MaxQuant msms.txt and evidence.txt.
+        """Reader for MaxQuant-like data (in terms of modification loading and decoy translation).
 
         See documentation of `PSMReaderBase` for more information.
 
@@ -178,6 +178,35 @@ def _translate_decoy(self) -> None:
                 self._psm_df[PsmDfCols.DECOY] == "-"
             ).astype(np.int8)
 
+    def _load_modifications(self, origin_df: pd.DataFrame) -> None:
+        if origin_df[self.mod_seq_column].str.contains("[", regex=False).any():
+            if origin_df[self.mod_seq_column].str.contains("(", regex=False).any():
+                origin_df[self.mod_seq_column] = origin_df[self.mod_seq_column].apply(
+                    replace_parentheses_with_brackets
+                )
+            mod_sep = "[]"
+        else:
+            mod_sep = "()"
+
+        seqs, mods, mod_sites = zip(
+            *origin_df[self.mod_seq_column].apply(
+                parse_mod_seq,
+                mod_sep=mod_sep,
+                fixed_C57=self.fixed_C57,
+            )
+        )
+
+        self._psm_df[PsmDfCols.MODS] = mods
+        self._psm_df[PsmDfCols.MOD_SITES] = mod_sites
+        if PsmDfCols.SEQUENCE not in self._psm_df.columns:
+            self._psm_df[PsmDfCols.SEQUENCE] = seqs
+
+
+class MaxQuantReader(ModifiedSequenceReader):
+    """Reader for MaxQuant data."""
+
+    _reader_type = "maxquant"
+
     def _pre_process(self, df: pd.DataFrame) -> pd.DataFrame:
         """MaxQuant-specific preprocessing of output data."""
         df = df[~pd.isna(df["Retention time"])]
@@ -201,26 +230,6 @@ def _pre_process(self, df: pd.DataFrame) -> pd.DataFrame:
         # min_rt = df['Retention time'].min()
         return df
 
-    def _load_modifications(self, origin_df: pd.DataFrame) -> None:
-        if origin_df[self.mod_seq_column].str.contains("[", regex=False).any():
-            if origin_df[self.mod_seq_column].str.contains("(", regex=False).any():
-                origin_df[self.mod_seq_column] = origin_df[self.mod_seq_column].apply(
-                    replace_parentheses_with_brackets
-                )
-            mod_sep = "[]"
-        else:
-            mod_sep = "()"
-
-        (seqs, self._psm_df[PsmDfCols.MODS], self._psm_df[PsmDfCols.MOD_SITES]) = zip(
-            *origin_df[self.mod_seq_column].apply(
-                parse_mod_seq,
-                mod_sep=mod_sep,
-                fixed_C57=self.fixed_C57,
-            )
-        )
-        if PsmDfCols.SEQUENCE not in self._psm_df.columns:
-            self._psm_df[PsmDfCols.SEQUENCE] = seqs
-
 
 def register_readers() -> None:
     """Register MaxQuant reader."""

alphabase/psm_reader/msfragger_reader.py

@@ -1,5 +1,6 @@
 """MSFragger reader."""
 
+import warnings
 from typing import List, Optional, Tuple
 
 import numpy as np
@@ -92,7 +93,7 @@ def __init__(
         raise NotImplementedError("MSFragger_PSM_TSV_Reader for psm.tsv")
 
 
-class MSFraggerPepXML(PSMReaderBase):
+class MSFraggerPepXMLReader(PSMReaderBase):
     """Reader for MSFragger's pep.xml file."""
 
     _reader_type = "msfragger_pepxml"
@@ -167,17 +168,10 @@ def _translate_decoy(self) -> None:
             self._psm_df[PsmDfCols.TO_REMOVE] += self._psm_df[PsmDfCols.DECOY] > 0
 
     def _translate_score(self) -> None:
-        # evalue score
+        """Translate MSFragger evalue to AlphaBase score: the larger the better."""
         self._psm_df[PsmDfCols.SCORE] = -np.log(self._psm_df[PsmDfCols.SCORE] + 1e-100)
 
     def _load_modifications(self, origin_df: pd.DataFrame) -> None:
-        if len(origin_df) == 0:
-            self._psm_df[PsmDfCols.MODS] = ""
-            self._psm_df[PsmDfCols.MOD_SITES] = ""
-            self._psm_df[PsmDfCols.AA_MASS_DIFFS] = ""
-            self._psm_df[PsmDfCols.AA_MASS_DIFF_SITES] = ""
-            return
-
         (
             self._psm_df[PsmDfCols.MODS],
             self._psm_df[PsmDfCols.MOD_SITES],
@@ -203,17 +197,30 @@ def _load_modifications(self, origin_df: pd.DataFrame) -> None:
                 inplace=True,
             )
 
-    def _post_process(self) -> None:
-        super()._post_process()
+    def _post_process(self, origin_df: pd.DataFrame) -> None:
         self._psm_df = (
             self._psm_df.query(f"{PsmDfCols.TO_REMOVE}==0")
             .drop(columns=PsmDfCols.TO_REMOVE)
             .reset_index(drop=True)
         )
+        super()._post_process(origin_df)
+
+
+class MSFraggerPepXML(MSFraggerPepXMLReader):
+    """Deprecated."""
+
+    def __init__(self, *args, **kwargs):
+        """Deprecated."""
+        warnings.warn(
+            "MSFraggerPepXML is deprecated and will ne removed in alphabase>1.5.0.",
+            "Please use the equivalent MSFraggerPepXMLReader instead.",
+            FutureWarning,
+        )
+        super().__init__(*args, **kwargs)
 
 
 def register_readers() -> None:
     """Register MSFragger readers."""
     psm_reader_provider.register_reader("msfragger_psm_tsv", MSFragger_PSM_TSV_Reader)
     psm_reader_provider.register_reader("msfragger", MSFragger_PSM_TSV_Reader)
-    psm_reader_provider.register_reader("msfragger_pepxml", MSFraggerPepXML)
+    psm_reader_provider.register_reader("msfragger_pepxml", MSFraggerPepXMLReader)

alphabase/psm_reader/pfind_reader.py

@@ -108,7 +108,7 @@ def _load_file(self, filename: str) -> pd.DataFrame:
     def _pre_process(self, df: pd.DataFrame) -> pd.DataFrame:
         """pFind-specific preprocessing of output data."""
         df.fillna("", inplace=True)
-        df = df[df.Sequence != ""]
+        df = df[df["Sequence"] != ""]
         df[PsmDfCols.RAW_NAME] = df["File_Name"].str.split(".").apply(lambda x: x[0])
         df["Proteins"] = df["Proteins"].apply(parse_pfind_protein)
         return df
@@ -119,23 +119,16 @@ def _translate_decoy(self) -> None:
         ).astype(np.int8)
 
     def _translate_score(self) -> None:
+        """Translate pFind pvalue to AlphaBase score: the larger the better."""
         self._psm_df[PsmDfCols.SCORE] = -np.log(
             self._psm_df[PsmDfCols.SCORE].astype(float) + 1e-100
         )
 
     def _load_modifications(self, origin_df: pd.DataFrame) -> None:
-        if len(origin_df) == 0:
-            self._psm_df[PsmDfCols.MODS] = ""
-            self._psm_df[PsmDfCols.MOD_SITES] = ""
-            return
+        mods, mod_sites = zip(*origin_df["Modification"].apply(get_pFind_mods))
 
-        (self._psm_df[PsmDfCols.MODS], self._psm_df[PsmDfCols.MOD_SITES]) = zip(
-            *origin_df["Modification"].apply(get_pFind_mods)
-        )
-
-        self._psm_df[PsmDfCols.MODS] = self._psm_df[PsmDfCols.MODS].apply(
-            translate_pFind_mod
-        )
+        self._psm_df[PsmDfCols.MODS] = [translate_pFind_mod(mod) for mod in mods]
+        self._psm_df[PsmDfCols.MOD_SITES] = mod_sites
 
 
 def register_readers() -> None: