Updated json reading to grouped jsons

uparma · Dec 5, 2018 · 7c3d79e · 7c3d79e
1 parent 2967edb
commit 7c3d79e
Show file tree

Hide file tree

Showing 4 changed files with 35 additions and 94 deletions.
diff --git a/examples/map_ursgal.py b/examples/map_ursgal.py
@@ -3,11 +3,13 @@
 import pprint
 
 def main():
-    up = uparma.UParma()
+    up = uparma.UParma(refresh_jsons=True)
     msgf_params = up.convert(
         {
             "precursor_mass_tolerance_unit": "ppm",
-            "min_pep_length" : 8
+            "min_pep_length" : 8,
+            "max_num_mods"  : 3,
+
         },
         target_style = 'msgfplus_style_1'
     )

diff --git a/uparma/parameters.json b/uparma/parameters.json
diff --git a/uparma/styles.json b/uparma/styles.json
@@ -1,30 +1 @@
-{
-    "msgfplus_style_1" : {
-        "name" : "MS-GF+",
-        "version" : [
-            "v2018.01.20",
-            "v2018.06.28",
-            "v2018.09.12"
-        ],
-        "reference" : "Kim S, Mischerikow N, Bandeira N, Navarro JD, Wich L, 'Mohammed S, Heck AJ, Pevzner PA. (2010) The Generating Function of CID, ETD, and CID/ETD Pairs of Tandem Mass Spectra: Applications to Database Search."
-    },
-    "msfragger_style_1" : {
-        "name" : "MSFragger",
-        "version" : [
-            "20170103"
-        ],
-        "reference" : "Kong, A. T., Leprevost, F. V, Avtonomov, D. M., Mellacheruvu, D., and Nesvizhskii, A. I. (2017) MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry-based proteomics. Nature Methods 14"
-    },
-    "ursgal_style_1" : {
-        "name" : "Ursgal",
-        "version" : [
-            "0.6.0",
-            "0.6.1"
-        ],
-        "reference" : "Kremer, L. P. M., Leufken, J., Oyunchimeg, P., Schulze, S. & Fufezan, C. (2016) Ursgal, Universal Python Module Combining Common Bottom-Up Proteomics Tools for Large-Scale Analysis. J. Proteome res. 15, 788-794."
-    },
-    "searchgui_style_1" : {},
-    "searchgui_style_2" : {},
-    "psi_ms_style_1" : {},
-    "xtandem_style_1" : {}
-}
+{'msgfplus_style_1': {'name': 'MS-GF+', 'version': ['v2018.01.30', 'v2018.06.28', 'v2018.09.12'], 'reference': "Kim S, Mischerikow N, Bandeira N, Navarro JD, Wich L, 'Mohammed S, Heck AJ, Pevzner PA. (2010) The Generating Function of CID, ETD, and CID/ETD Pairs of Tandem Mass Spectra: Applications to Database Search."}, 'msfragger_style_1': {'name': 'MSFragger', 'version': ['20170103'], 'reference': 'Kong, A. T., Leprevost, F. V, Avtonomov, D. M., Mellacheruvu, D., and Nesvizhskii, A. I. (2017) MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry-based proteomics. Nature Methods 14'}, 'ursgal_style_1': {'name': 'Ursgal', 'version': ['0.6.0', '0.6.1'], 'reference': 'Kremer, L. P. M., Leufken, J., Oyunchimeg, P., Schulze, S. & Fufezan, C. (2016) Ursgal, Universal Python Module Combining Common Bottom-Up Proteomics Tools for Large-Scale Analysis. J. Proteome res. 15, 788-794.'}, 'searchgui_style_1': {}, 'searchgui_style_2': {}, 'psi_ms_style_1': {}, 'xtandem_style_1': {'name': 'X!Tandem', 'version': ['vengeance', 'alanine'], 'reference': ''}, 'masamanda_style_1': {'name': 'MS Amanda', 'version': ['2_0_0_9695', '2_0_0_9706', '2_0_0_10695']}}
diff --git a/uparma/uparma.py b/uparma/uparma.py
@@ -1,11 +1,14 @@
+#!/usr/bin/env python3
 import requests
 import os
 import json
 from collections import defaultdict as ddict
 
 URLS = {
-    'parameter_url' : 'https://raw.githubusercontent.com/uparma/uparma-lib/master/parameters.json',
-    'styles_url' : 'https://raw.githubusercontent.com/uparma/uparma-lib/master/styles.json'
+    ('general', 'parameters') : 'https://raw.githubusercontent.com/uparma/uparma-lib/master/general_parameters.json',
+    ('spectrum', 'parameters') : 'https://raw.githubusercontent.com/uparma/uparma-lib/master/spectrum_parameters.json',
+    ('modifications', 'parameters') : 'https://raw.githubusercontent.com/uparma/uparma-lib/master/modification_parameters.json',
+    ('general', 'styles') : 'https://raw.githubusercontent.com/uparma/uparma-lib/master/styles.json'
 }
 
 
@@ -44,12 +47,11 @@ def __init__(self, refresh_jsons=False, source_style="ursgal_style_1"):
                 with requests.get(url) as req:
                     with open(json_file_name, 'w') as j:
                         print(req.json(), file=j)
-                    self.jsons[json_file_name] = req.json()
+                    self.jsons[url_id] = req.json()
             else:
                 for url_id, url in URLS.items():
-
                     with open(full_path) as j:
-                        self.jsons[json_file_name] = json.load(j)
+                        self.jsons[url_id] = json.load(j)
 
         self._parse_jsons()
 
@@ -63,24 +65,27 @@ def _parse_jsons(self):
             }
 
         """
-        for uparma_entry in self.jsons['parameters.json']:
-
-            _id = uparma_entry['_id']
-            self.parameters[_id] = uparma_entry
-
-            for key, value in uparma_entry.items():
-                if "style" in key:
-                    if isinstance(value, list):
-                        value = ", ".join(value)
-                    try:
-                        self.parameter2id[key][value] = _id
-                    except:
-                        self.parameter2id[key] = {value: _id}
-                        self.available_styles.append(key)
-
-            assert _id in self.parameters.keys(), """
-            ID {0} is not unique in parameters.json
-            """.format(_id)
+        for url_id in self.jsons.keys():
+            json_tag, json_type = url_id
+            if json_type != 'parameters':
+                continue
+            for uparma_entry in self.jsons[url_id]:
+                _id = uparma_entry['_id']
+                self.parameters[_id] = uparma_entry
+
+                for key, value in uparma_entry.items():
+                    if "style" in key:
+                        if isinstance(value, list):
+                            value = ", ".join(value)
+                        try:
+                            self.parameter2id[key][value] = _id
+                        except:
+                            self.parameter2id[key] = {value: _id}
+                            self.available_styles.append(key)
+
+                assert _id in self.parameters.keys(), """
+                ID {0} is not unique in parameters.json
+                """.format(_id)
 
     def convert(self, param_dict, target_style=None):
         """
@@ -153,4 +158,6 @@ def translate(self, param_dict, source_style=None, target_style=None):
 
 
 
+if __name__ == '__main__':
+    UParma(refresh_jsons=True)