Merge pull request #177 from iomega/implement_onnx

Define zenodo doi

CHANGELOG.md

@@ -5,6 +5,9 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.7.1]
+- Define the newest zenodo DOI in one location. 
+
 ## [0.7.0]
 - Store random forest model in Onnx format
 

README.md

@@ -62,8 +62,8 @@ ms2query --library .\folder_to_store_the_library --download --ionmode positive
 ```
 
 Alternatively all model files can be manually downloaded from 
-https://zenodo.org/record/6997924#.YvuonS5BxPY for positive mode and 
-https://zenodo.org/record/7107654#.Yy3BeKRBxPY for negative mode.
+https://zenodo.org/record/7753249#.ZBmO_sLMJPY for positive mode and 
+https://zenodo.org/record/7753267#.ZBmPYsLMJPY for negative mode.
 
 #### Preprocessing mass spectra
 MS2Query is run on all MS2 spectra in a spectrum file. 
@@ -136,11 +136,8 @@ ms2query_library_files_directory = "./ms2query_library_files"
 ms2_spectra_directory = 
 ion_mode = # Fill in "positive" or "negative" to indicate for which ion mode you would like to download the library
 
-zenodo_DOIs = {"positive": 7753249, 
-               "negative": 7753267}
-
 # Downloads pretrained models and files for MS2Query (>2GB download)
-download_zenodo_files(zenodo_DOIs[ion_mode], 
+download_zenodo_files(ion_mode, 
                       ms2query_library_files_directory)
 
 # Create a MS2Library object

ms2query/__init__.py

@@ -70,9 +70,7 @@ def command_line():
                                    additional_metadata_columns=additional_columns)
     if args.download:
         assert ion_mode is not None, "Ion mode should be specified by adding --ion_mode"
-        zenodo_DOIs = {"positive": 7753249,
-                       "negative": 7753267}
-        download_zenodo_files(zenodo_DOIs[ion_mode],
+        download_zenodo_files(ion_mode,
                               ms2query_library_files_directory)
 
     if ms2_spectra_location is not None:

ms2query/__version__.py

@@ -1 +1 @@
-__version__ = '0.7.0'
+__version__ = '0.7.1'

ms2query/run_ms2query.py

@@ -6,8 +6,16 @@
 from ms2query.utils import load_matchms_spectrum_objects_from_file, SettingsRunMS2Query, return_non_existing_file_name
 
 
-def download_zenodo_files(zenodo_doi: int,
-                          dir_to_store_files:str):
+def zenodo_dois(ionisation_mode):
+    "Returns the most up to date DOI for Zenodo"
+    zenodo_DOIs = {"positive": 7753249,
+                   "negative": 7753267}
+    assert ionisation_mode in zenodo_DOIs, "Expected 'positive' or 'negative' as input"
+    return zenodo_DOIs[ionisation_mode]
+
+
+def download_zenodo_files(ionisation_mode: str,
+                          dir_to_store_files: str):
     """Downloads files from Zenodo
 
     Args:
@@ -20,6 +28,7 @@ def download_zenodo_files(zenodo_doi: int,
         """
     if not os.path.exists(dir_to_store_files):
         os.mkdir(dir_to_store_files)
+    zenodo_doi = zenodo_dois(ionisation_mode)
     file_names_metadata_url = "https://zenodo.org/api/records/" + str(zenodo_doi)
     with urlopen(file_names_metadata_url) as zenodo_metadata_file:
         file_names_metadata_json: dict = json.loads(zenodo_metadata_file.read())

tests/test_run_ms2query.py

@@ -17,25 +17,13 @@ def test_download_default_models(tmp_path):
     """Tests downloading small files from zenodo
 
     The files are a total of 20 MB from https://zenodo.org/record/7108049#.Yy2nPKRBxPY"""
-    dir_to_store_test_files = os.path.join(tmp_path, "positive_model")
-    download_zenodo_files(7108049, dir_to_store_test_files)
-    assert os.path.exists(dir_to_store_test_files)
-    assert os.path.exists(os.path.join(dir_to_store_test_files,
-                                       "GNPS_15_12_2021_neg_test_250_inchikeys.pickle"))
-    assert os.path.exists(os.path.join(dir_to_store_test_files,
-                                       "GNPS_15_12_2021_neg_test_3000_spectra.pickle"))
-    assert os.path.exists(os.path.join(dir_to_store_test_files,
-                                       "GNPS_15_12_2021_pos_test_250_inchikeys.pickle"))
-    assert os.path.exists(os.path.join(dir_to_store_test_files,
-                                       "GNPS_15_12_2021_pos_test_3000_spectra.pickle"))
-
     run_test = False # Run test is set to false, since downloading takes too long for default testing
     if run_test:
         dir_to_store_positive_files = os.path.join(tmp_path, "positive_model")
         dir_to_store_negative_files = os.path.join(tmp_path, "negative_model")
 
-        download_zenodo_files(6997924, dir_to_store_positive_files)
-        download_zenodo_files(7107654, dir_to_store_negative_files)
+        download_zenodo_files("positive", dir_to_store_positive_files)
+        download_zenodo_files("negative", dir_to_store_negative_files)
         assert os.path.exists(dir_to_store_positive_files)
         assert os.path.exists(dir_to_store_negative_files)
         pos_ms2library = create_library_object_from_one_dir(dir_to_store_positive_files)