Merge branch 'development' into feature/ptm_model_support

.cookietemple.yml

@@ -15,5 +15,5 @@ full_name: Victor Giurcoiu
 email: victor.giurcoiu@tum.de
 project_name: spectrum_fundamentals
 project_short_description: Fundamentals public repo
-version: 0.5.0
+version: 0.5.1
 license: MIT

.github/release-drafter.yml

@@ -1,5 +1,5 @@
-name-template: "0.5.0 🌈" # <<COOKIETEMPLE_FORCE_BUMP>>
-tag-template: 0.5.0 # <<COOKIETEMPLE_FORCE_BUMP>>
+name-template: "0.5.1 🌈" # <<COOKIETEMPLE_FORCE_BUMP>>
+tag-template: 0.5.1 # <<COOKIETEMPLE_FORCE_BUMP>>
 exclude-labels:
     - "skip-changelog"
 

.github/workflows/labeler.yml

@@ -14,6 +14,6 @@ jobs:
               uses: actions/checkout@v2.3.3
 
             - name: Run Labeler
-              uses: crazy-max/ghaction-github-labeler@v3.1.1
+              uses: crazy-max/ghaction-github-labeler@v5.0.0
               with:
                   skip-delete: true

.github/workflows/main_master_branch_protection.yml

@@ -20,7 +20,7 @@ jobs:
             # NOTE - this may not work if the PR is coming from a fork, due to limitations in GitHub actions secrets
             - name: Post PR comment
               if: failure()
-              uses: mshick/add-pr-comment@v1
+              uses: mshick/add-pr-comment@v2
               with:
                   message: |
                       Hi @${{ github.event.pull_request.user.login }},
@@ -61,7 +61,7 @@ jobs:
             # NOTE - this may not work if the PR is coming from a fork, due to limitations in GitHub actions secrets
             - name: Post PR comment
               if: failure()
-              uses: mshick/add-pr-comment@v1
+              uses: mshick/add-pr-comment@v2
               with:
                   message: |
                       Hi @${{ github.event.pull_request.user.login }},

.github/workflows/run_tests.yml

@@ -104,7 +104,7 @@ jobs:
                   print("::set-output name=result::{}".format(result))
 
             - name: Restore pre-commit cache
-              uses: actions/cache@v2.1.7
+              uses: actions/cache@v4.0.2
               if: matrix.session == 'pre-commit'
               with:
                   path: ~/.cache/pre-commit

cookietemple.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.5.0
+current_version = 0.5.1
 
 [bumpversion_files_whitelisted]
 init_file = spectrum_fundamentals/__init__.py

docs/conf.py

@@ -54,9 +54,9 @@
 # the built documents.
 #
 # The short X.Y version.
-version = "0.5.0"
+version = "0.5.1"
 # The full version, including alpha/beta/rc tags.
-release = "0.5.0"
+release = "0.5.1"
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

noxfile.py

@@ -1,4 +1,5 @@
 """Nox sessions."""
+
 import os
 import shlex
 import shutil

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "spectrum_fundamentals"
-version = "0.5.0"  # <<COOKIETEMPLE_FORCE_BUMP>>
+version = "0.5.1"  # <<COOKIETEMPLE_FORCE_BUMP>>
 description = "Fundamental functions, annotation pipeline and constants for oktoberfest"
 authors = ["Wilhelmlab at Technical University of Munich"]
 license = "MIT"

requirements.txt

@@ -1,8 +1,29 @@
-click>=8.0.0
-rich>=10.3.0
-PyYAML>=5.4.1
-numpy>=1.24.1
-pandas>=1.3.0
-scikit-learn>=1.0
-joblib>=1.0.1
-moepy>=1.1.4
+click==8.1.7 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0"
+colorama==0.4.6 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0" and platform_system == "Windows"
+contourpy==1.1.1 ; python_version >= "3.8" and python_full_version < "3.11.0"
+cycler==0.12.1 ; python_version >= "3.8" and python_full_version < "3.11.0"
+fonttools==4.51.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+importlib-resources==6.4.0 ; python_version >= "3.8" and python_version < "3.10"
+joblib==1.4.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+kiwisolver==1.4.5 ; python_version >= "3.8" and python_full_version < "3.11.0"
+markdown-it-py==3.0.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+matplotlib==3.7.5 ; python_version >= "3.8" and python_full_version < "3.11.0"
+mdurl==0.1.2 ; python_version >= "3.8" and python_full_version < "3.11.0"
+moepy==1.1.4 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0"
+numpy==1.24.4 ; python_version >= "3.8" and python_full_version < "3.11.0"
+packaging==24.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+pandas==1.5.3 ; python_version >= "3.8" and python_full_version < "3.11.0"
+pillow==10.3.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+pygments==2.17.2 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0"
+pyparsing==3.1.2 ; python_version >= "3.8" and python_full_version < "3.11.0"
+python-dateutil==2.9.0.post0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+pytz==2024.1 ; python_version >= "3.8" and python_full_version < "3.11.0"
+pyyaml==6.0.1 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0"
+rich==13.7.1 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0"
+scikit-learn==1.3.2 ; python_version >= "3.8" and python_full_version < "3.11.0"
+scipy==1.10.1 ; python_version >= "3.8" and python_full_version < "3.11.0"
+six==1.16.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+threadpoolctl==3.4.0 ; python_version >= "3.8" and python_full_version < "3.11.0"
+tqdm==4.66.2 ; python_full_version >= "3.8.0" and python_full_version < "3.11.0"
+typing-extensions==4.11.0 ; python_version >= "3.8" and python_version < "3.9"
+zipp==3.18.1 ; python_version >= "3.8" and python_version < "3.10"

spectrum_fundamentals/__init__.py

@@ -1,7 +1,8 @@
 """Initialize fundamentals."""
+
 __author__ = "Mario Picciani"
 __email__ = "mario.picciani@tum.de"
-__version__ = "0.5.0"
+__version__ = "0.5.1"
 
 import logging
 import logging.handlers

spectrum_fundamentals/__main__.py

@@ -5,7 +5,7 @@
 
 
 @click.command()
-@click.version_option(version="0.5.0", message=click.style("spectrum_fundamentals Version: 0.5.0"))
+@click.version_option(version="0.5.1", message=click.style("spectrum_fundamentals Version: 0.5.1"))
 def main() -> None:
     """spectrum_fundamentals."""
 

spectrum_fundamentals/charge.py

@@ -22,7 +22,7 @@ def indices_to_one_hot(labels: Union[int, List[int], np.ndarray], classes: Optio
             f"Type of labels not understood. Only int, List[int] and np.ndarray are supported. Given: {type(labels)}."
         )
 
-    max_label = labels.max()
+    max_label = int(labels.max())
     if classes is None:
         classes = max_label
     if max_label > classes:

spectrum_fundamentals/constants.py

@@ -109,6 +109,17 @@
     "n[214]": "[UNIMOD:214]-",
 }
 
+#######################
+# Xisearch constants #
+#######################
+
+XISEARCH_VAR_MODS = {
+    "ox": "[UNIMOD:35]",
+    "cm": "[UNIMOD:4]",
+    "dsso": "[UNIMOD:1896]",
+    "dsbu": "[UNIMOD:1884]",
+}
+
 ####################
 # MASS CALCULATION #
 ####################
@@ -162,51 +173,49 @@
     "[UNIMOD:730]": 304.205360,  # iTRAQ8
     "[UNIMOD:259]": 8.014199,  # SILAC Lysine
     "[UNIMOD:267]": 10.008269,  # SILAC Arginine
-
-    "[]":0.0,
-    "[UNIMOD:1]" :42.010565, #Acetylation
-    "[UNIMOD:122]" :27.994915, #Formylation
-    "[UNIMOD:1289]" :70.041865, #Butyrylation
-    "[UNIMOD:1363]" :68.026215, #Crotonylation
-    "[UNIMOD:1848]" :114.031694, #Glutarylation
-    "[UNIMOD:1914]" :-32.008456, #Oxidation and then loss of oxidized M side chain
-    "[UNIMOD:2]" :-0.984016, #Amidation
-    "[UNIMOD:21]" :79.966331, #Phosphorylation
-    "[UNIMOD:213]" :541.06111, #ADP-ribosylation
-    "[UNIMOD:23]" :-18.010565, #Water Loss
-    "[UNIMOD:24]" :71.037114, #Propionamidation
-    "[UNIMOD:354]" :44.985078, #Nitrosylation
-    "[UNIMOD:28]" :-17.026549, #Glu to PyroGlu
-    "[UNIMOD:280]" :28.0313, #Ethylation
-    "[UNIMOD:299]" :43.989829, #Carboxylation
-    "[UNIMOD:3]" :226.077598, #Biotinylation
-    "[UNIMOD:34]" :14.01565, #Methylation
-    "[UNIMOD:345]" :47.984744, #Trioxidation
-    "[UNIMOD:35]" :15.994915, #Hydroxylation
-    "[UNIMOD:35]" :15.994915, #Oxidation
-    "[UNIMOD:351]" :3.994915, #Oxidation to Kynurenine
-    "[UNIMOD:36]" :28.0313, #Dimethylation
-    "[UNIMOD:360]" :-30.010565, #Pyrrolidinone
-    "[UNIMOD:368]" :-33.987721, #Dehydroalanine
-    "[UNIMOD:37]" :42.04695, #Trimethylation
-    "[UNIMOD:385]" :-17.026549, #Ammonia loss
-    "[UNIMOD:392]" :29.974179, #Quinone
-    "[UNIMOD:4]" :57.021464, #Carbamidomethyl
-    "[UNIMOD:40]" :79.956815, #Sulfonation
-    "[UNIMOD:401]" :-2.01565, #Didehydro
-    "[UNIMOD:425]" :31.989829, #Dioxidation
-    "[UNIMOD:43]" :203.079373, #HexNAc
-    "[UNIMOD:44]" :204.187801, #Farnesylation
-    "[UNIMOD:447]" :-15.994915, #Reduction
-    "[UNIMOD:46]" :229.014009, #Pyridoxal phosphate
-    "[UNIMOD:47]" :238.229666, #Palmitoylation
-    "[UNIMOD:5]" :43.005814, #Carbamyl
-    "[UNIMOD:58]" :56.026215, #Propionylation
-    "[UNIMOD:6]" :58.005479, #Carboxymethylation
-    "[UNIMOD:64]" :100.016044, #Succinylation
-    "[UNIMOD:7]" :0.984016, #Deamidation
-    "[UNIMOD:747]" :86.000394, #Malonylation
-
+    "[]": 0.0,
+    "[UNIMOD:1]": 42.010565,  # Acetylation
+    "[UNIMOD:122]": 27.994915,  # Formylation
+    "[UNIMOD:1289]": 70.041865,  # Butyrylation
+    "[UNIMOD:1363]": 68.026215,  # Crotonylation
+    "[UNIMOD:1848]": 114.031694,  # Glutarylation
+    "[UNIMOD:1914]": -32.008456,  # Oxidation and then loss of oxidized M side chain
+    "[UNIMOD:2]": -0.984016,  # Amidation
+    "[UNIMOD:21]": 79.966331,  # Phosphorylation
+    "[UNIMOD:213]": 541.06111,  # ADP-ribosylation
+    "[UNIMOD:23]": -18.010565,  # Water Loss
+    "[UNIMOD:24]": 71.037114,  # Propionamidation
+    "[UNIMOD:354]": 44.985078,  # Nitrosylation
+    "[UNIMOD:28]": -17.026549,  # Glu to PyroGlu
+    "[UNIMOD:280]": 28.0313,  # Ethylation
+    "[UNIMOD:299]": 43.989829,  # Carboxylation
+    "[UNIMOD:3]": 226.077598,  # Biotinylation
+    "[UNIMOD:34]": 14.01565,  # Methylation
+    "[UNIMOD:345]": 47.984744,  # Trioxidation
+    "[UNIMOD:35]": 15.994915,  # Hydroxylation
+    "[UNIMOD:35]": 15.994915,  # Oxidation
+    "[UNIMOD:351]": 3.994915,  # Oxidation to Kynurenine
+    "[UNIMOD:36]": 28.0313,  # Dimethylation
+    "[UNIMOD:360]": -30.010565,  # Pyrrolidinone
+    "[UNIMOD:368]": -33.987721,  # Dehydroalanine
+    "[UNIMOD:37]": 42.04695,  # Trimethylation
+    "[UNIMOD:385]": -17.026549,  # Ammonia loss
+    "[UNIMOD:392]": 29.974179,  # Quinone
+    "[UNIMOD:4]": 57.021464,  # Carbamidomethyl
+    "[UNIMOD:40]": 79.956815,  # Sulfonation
+    "[UNIMOD:401]": -2.01565,  # Didehydro
+    "[UNIMOD:425]": 31.989829,  # Dioxidation
+    "[UNIMOD:43]": 203.079373,  # HexNAc
+    "[UNIMOD:44]": 204.187801,  # Farnesylation
+    "[UNIMOD:447]": -15.994915,  # Reduction
+    "[UNIMOD:46]": 229.014009,  # Pyridoxal phosphate
+    "[UNIMOD:47]": 238.229666,  # Palmitoylation
+    "[UNIMOD:5]": 43.005814,  # Carbamyl
+    "[UNIMOD:58]": 56.026215,  # Propionylation
+    "[UNIMOD:6]": 58.005479,  # Carboxymethylation
+    "[UNIMOD:64]": 100.016044,  # Succinylation
+    "[UNIMOD:7]": 0.984016,  # Deamidation
+    "[UNIMOD:747]": 86.000394,  # Malonylation
     #
 }
 MOD_MASSES_SAGE = {

spectrum_fundamentals/metrics/fragments_ratio.py

@@ -31,7 +31,7 @@ class FragmentsRatio(Metric):
 
     @staticmethod
     def count_with_ion_mask(
-        boolean_array: scipy.sparse.csr_matrix, ion_mask: Optional[Union[np.ndarray, scipy.sparse.csr_matrix]] = None
+        boolean_array: scipy.sparse.csr_matrix, ion_mask: Optional[Union[np.ndarray, scipy.sparse.spmatrix]] = None
     ) -> np.ndarray:
         """
         Count the number of ions.
@@ -43,8 +43,6 @@ def count_with_ion_mask(
         :return: number of observed/predicted peaks not masked by ion_mask
         """
         if ion_mask is None:
-            ion_mask = []
-        if len(ion_mask) == 0:
             ion_mask = scipy.sparse.csr_matrix(np.ones((174, 1)))
         else:
             ion_mask = scipy.sparse.csr_matrix(ion_mask).T