ENH: Added mzML formats and type

.pre-commit-config.yaml

@@ -25,6 +25,8 @@ repos:
     rev: 4.0.1
     hooks:
       - id: flake8
+        args:
+          - --max-line-length=88
         additional_dependencies:
           - pycodestyle==2.8.0
           - pyflakes==2.4.0

conda-recipe/meta.yaml

@@ -20,6 +20,7 @@ requirements:
     - versioningit
 
   run:
+    - pymzml
     - qiime2 {{ qiime2_epoch }}.*
     - q2-types {{ qiime2_epoch }}.*
     - q2templates {{ qiime2_epoch }}.*

q2_ms/__init__.py

@@ -1,13 +1,15 @@
-# flake8: noqa
 # ----------------------------------------------------------------------------
 # Copyright (c) 2024, QIIME 2 development team.
 #
 # Distributed under the terms of the Modified BSD License.
 #
 # The full license is in the file LICENSE, distributed with this software.
 # ----------------------------------------------------------------------------
+import importlib
 
 try:
     from ._version import __version__
 except ModuleNotFoundError:
     __version__ = "0.0.0+notfound"
+
+importlib.import_module("q2_ms.types")

q2_ms/citations.bib

@@ -0,0 +1,10 @@
+@article{kosters2018pymzml,
+  title={pymzML v2. 0: introducing a highly compressed and seekable gzip format},
+  author={K{\"o}sters, M and Leufken, Johannes and Schulze, Stefan and Sugimoto, K and Klein, Joshua and Zahedi, RP and Hippler, Michael and Leidel, SA and Fufezan, Christian},
+  journal={Bioinformatics},
+  volume={34},
+  number={14},
+  pages={2513--2514},
+  year={2018},
+  publisher={Oxford University Press}
+}

q2_ms/plugin_setup.py

@@ -5,10 +5,11 @@
 #
 # The full license is in the file LICENSE, distributed with this software.
 # ----------------------------------------------------------------------------
-
+from q2_types.sample_data import SampleData
 from qiime2.plugin import Citations, Plugin
 
 from q2_ms import __version__
+from q2_ms.types import mzML, mzMLDirFmt, mzMLFormat
 
 citations = Citations.load("citations.bib", package="q2_ms")
 
@@ -19,5 +20,16 @@
     package="q2_ms",
     description="A QIIME 2 plugin for MS data processing.",
     short_description="A QIIME 2 plugin for MS data processing.",
-    citations=[],
+)
+
+# Registrations
+plugin.register_semantic_types(
+    mzML,
+)
+
+plugin.register_semantic_type_to_format(SampleData[mzML], artifact_format=mzMLDirFmt)
+
+plugin.register_formats(
+    mzMLFormat,
+    mzMLDirFmt,
 )

q2_ms/tests/__init__.py → q2_ms/types/__init__.py

@@ -1,8 +1,11 @@
-# flake8: noqa
 # ----------------------------------------------------------------------------
 # Copyright (c) 2024, QIIME 2 development team.
 #
 # Distributed under the terms of the Modified BSD License.
 #
 # The full license is in the file LICENSE, distributed with this software.
 # ----------------------------------------------------------------------------
+from q2_ms.types._format import mzMLDirFmt, mzMLFormat
+from q2_ms.types._type import mzML
+
+__all__ = ["mzMLFormat", "mzMLDirFmt", "mzML"]

q2_ms/types/_format.py

@@ -0,0 +1,37 @@
+# ----------------------------------------------------------------------------
+# Copyright (c) 2024, QIIME 2 development team.
+#
+# Distributed under the terms of the Modified BSD License.
+#
+# The full license is in the file LICENSE, distributed with this software.
+# ----------------------------------------------------------------------------
+import os
+import sys
+
+import pymzml
+from qiime2.core.exceptions import ValidationError
+from qiime2.plugin import model
+
+
+class mzMLFormat(model.TextFileFormat):
+    def _validate(self, n_records=None):
+        try:
+            # Suppressing warning print "Not index found and build_index_from_scratch
+            # is False". This could also be solved with setting build_index_from_scratch
+            # to True but this builds the index and slows down validation.
+            sys.stdout = open(os.devnull, "w")
+            pymzml.run.Reader(str(self))
+            sys.stdout = sys.__stdout__
+        except Exception as e:
+            raise ValidationError(e)
+
+    def _validate_(self, level):
+        self._validate()
+
+
+class mzMLDirFmt(model.DirectoryFormat):
+    mzml = model.FileCollection(r".*\.mzML$", format=mzMLFormat)
+
+    @mzml.set_path_maker
+    def mzml_path_maker(self, sample_id):
+        return f"{sample_id}.mzML"

q2_ms/tests/test_methods.py → q2_ms/types/_type.py

@@ -5,11 +5,7 @@
 #
 # The full license is in the file LICENSE, distributed with this software.
 # ----------------------------------------------------------------------------
-from qiime2.plugin.testing import TestPluginBase
+from q2_types.sample_data import SampleData
+from qiime2.core.type import SemanticType
 
-
-class Test(TestPluginBase):
-    package = "q2_ms.tests"
-
-    def test(self):
-        pass
+mzML = SemanticType("mzML", variant_of=SampleData.field["type"])

q2_ms/types/tests/data/mzML_invalid/invalid.mzML

@@ -0,0 +1,15 @@
+<?xml version="1.0" encoding="ISO-8859-1"?>
+<indexedmzML xmlns="http://psi.hupo.org/ms/mzml" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://psi.hupo.org/ms/mzml http://psidev.info/files/ms/mzML/xsd/mzML1.1.0_idx.xsd">
+  <mzML xmlns="http://psi.hupo.org/ms/mzml" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://psi.hupo.org/ms/mzml http://psidev.info/files/ms/mzML/xsd/mzML1.1.0.xsd" id="urn:lsid:psidev.info:mzML.instanceDocuments.tiny.pwiz" version="1.1.0">
+    <cvList count="2">
+      <cv id="MS" fullName="Proteomics Standards Initiative Mass Spectrometry Ontology" version="2.26.0" URI="http://psidev.cvs.sourceforge.net/*checkout*/psidev/psi/psi-ms/mzML/controlledVocabulary/psi-ms.obo"/>
+      <cv id="UO" fullName="Unit Ontology" version="14:07:2009" URI="http://obo.cvs.sourceforge.net/*checkout*/obo/obo/ontology/phenotype/unit.obo"/>
+    </cvList>
+    <fileDescription>
+      <fileContent>
+        <cvParam cvRef="MS" accession="MS:1000580" name="MSn spectrum" value=""/>
+        <cvParam cvRef="MS" accession="MS:1000127" name="centroid spectrum" value=""/>
+      </fileContent>
+      <sourceFileList count="3">
+        <sourceFile id="tiny1.yep" name="tiny1.yep" location="file://F:/data/Exp01">
+          <cvParam cvRef="MS" accession="MS:1000567" name="Bruker/Agilent YEP file" value=""/>