Add ProLuCID parquet support

README.rst

@@ -99,6 +99,7 @@ Supported file formats
  `Percolator tab <https://github.com/percolator/percolator/wiki/Interface>`_                                           ``percolator``           ✅              ✅
  Proteome Discoverer MSF                                                                                               ``proteome_discoverer``  ✅              ❌
  `Sage <https://github.com/lazear/sage/blob/v0.12.0/DOCS.md#interpreting-sage-output>`_                                ``sage``                 ✅              ❌
+ TIMScore Parquet                                                                                                      ``timscore``             ✅              ❌
  `TSV <https://psm-utils.readthedocs.io/en/stable/api/psm_utils.io/#module-psm_utils.io.tsv>`_                         ``tsv``                  ✅              ✅
  `X!Tandem XML <https://www.thegpm.org/tandem/>`_                                                                      ``xtandem``              ✅              ❌
 ===================================================================================================================== ======================== =============== ===============

docs/source/api/psm_utils.io.rst

@@ -104,6 +104,14 @@ psm_utils.io.tsv
 
 
 
+psm_utils.io.timscore
+################################
+.. automodule:: psm_utils.io.timscore
+   :members:
+   :inherited-members:
+
+
+
 psm_utils.io.xtandem
 ####################
 

psm_utils/io/__init__.py

@@ -18,6 +18,7 @@
 import psm_utils.io.percolator as percolator
 import psm_utils.io.proteome_discoverer as proteome_discoverer
 import psm_utils.io.sage as sage
+import psm_utils.io.timscore as timscore
 import psm_utils.io.tsv as tsv
 import psm_utils.io.xtandem as xtandem
 from psm_utils.io._base_classes import WriterBase
@@ -68,6 +69,12 @@
         "extension": ".msf",
         "filename_pattern": r"^.*\.msf$",
     },
+    "timscore": {
+        "reader": timscore.TIMScoreReader,
+        "writer": None,
+        "extension": ".parquet",
+        "filename_pattern": r"^.*\.candidates\.parquet$",
+    },
     "tsv": {
         "reader": tsv.TSVReader,
         "writer": tsv.TSVWriter,

psm_utils/io/timscore.py

@@ -0,0 +1,89 @@
+"""Reader for Proteome Discoverer MSF PSM files."""
+
+import logging
+from pathlib import Path
+from typing import Union
+
+import numpy as np
+import pandas as pd
+
+from psm_utils import PSM
+from psm_utils.io._base_classes import ReaderBase
+
+logger = logging.getLogger(__name__)
+
+
+class TIMScoreReader(ReaderBase):
+    """Reader for TIMScore Parquet files."""
+
+    def __init__(
+        self,
+        filename: Union[str, Path],
+        *args,
+        **kwargs,
+    ) -> None:
+        """
+        Reader for TIMScore Parquet files.
+
+        Parameters
+        ----------
+        filename: str, pathlib.Path
+            Path to MSF file.
+
+        """
+        super().__init__(filename, *args, **kwargs)
+
+        self.data = pd.read_parquet(self.filename)
+
+    def __len__(self):
+        """Return number of PSMs in file."""
+        return len(self.data)
+
+    def __iter__(self):
+        """Iterate over file and return PSMs one-by-one."""
+        for entry in self.data.itertuples():
+            yield PSM(
+                peptidoform=_parse_peptidoform(
+                    entry.stripped_peptide, entry.ptms, entry.ptm_locations, entry.precursor_charge
+                ),
+                spectrum_id=entry.ms2_id,
+                is_decoy=None,  # TODO: Parse from protein?
+                score=entry.tims_score,  # TODO: Correct score?
+                precursor_mz=entry.precursor_mz,
+                retention_time=entry.rt,
+                ion_mobility=entry.corrected_ook0,
+                protein_list=list(entry.locus_name),
+                rank=entry.rank,
+                source="TIMScore",
+                provenance_data={
+                    "candidate_id": str(entry.candidate_id),
+                    "ms2_id": str(entry.ms2_id),
+                    "parent_id": str(entry.parent_id),
+                },
+                metadata={
+                    "leading_aa": str(entry.leading_aa),
+                    "trailing_aa": str(entry.trailing_aa),
+                    "ook0": str(entry.ook0),
+                },
+                rescoring_features={
+                    "x_corr_score": float(entry.x_corr_score),
+                    "delta_cn_score": float(entry.delta_cn_score),
+                    "ppm_error": float(entry.ppm_error),
+                    "number_matched_ions": float(entry.number_matched_ions),
+                    "number_expected_ions": float(entry.number_expected_ions),
+                    "ion_proportion": float(entry.ion_proportion),
+                    "spectrum_total_ion_intensity": float(entry.spectrum_total_ion_intensity),
+                },
+            )
+
+
+def _parse_peptidoform(
+    stripped_peptide: str, ptms: np.ndarray, ptm_locations: np.ndarray, precursor_charge: int
+) -> str:
+    """Parse peptide sequence and modifications to ProForma."""
+    # TODO: How are terminal modifications handled?
+    peptidoform = list(stripped_peptide)
+    for ptm, ptm_location in zip(ptms, ptm_locations):
+        peptidoform[ptm_location] = f"{peptidoform[ptm_location]}[{ptm}]"
+    peptidoform.append(f"/{precursor_charge}")
+    return "".join(peptidoform)

tests/test_io/test_timscore.py

@@ -0,0 +1,14 @@
+import numpy as np
+
+from psm_utils.io.timscore import _parse_peptidoform
+
+
+def test_parse_peptidoform():
+    test_cases = [
+        ("ACDMEK", np.array([]), np.array([]), 2, "ACDMEK/2"),
+        ("ACDMEK", np.array([15.99]), np.array([3]), 2, "ACDM[15.99]EK/2"),
+        ("ACDMEK", np.array([57.02, 15.99]), np.array([1, 3]), 2, "AC[57.02]DM[15.99]EK/2"),
+    ]
+
+    for peptide, ptms, ptm_locations, precursor_charge, expected in test_cases:
+        assert _parse_peptidoform(peptide, ptms, ptm_locations, precursor_charge) == expected