update constant docs

PathwayMerger · Mar 12, 2020 · 585fc00 · 585fc00
1 parent 24f29e4
commit 585fc00
Show file tree

Hide file tree

Showing 2 changed files with 49 additions and 15 deletions.
diff --git a/src/pathme/constants.py b/src/pathme/constants.py
@@ -1,6 +1,9 @@
 # -*- coding: utf-8 -*-
 
-"""This module contains all the constants used in PathMe repo."""
+"""Constants for PathMe.
+
+This module contains all the string constants used in PathMe.
+"""
 
 import os
 
@@ -21,17 +24,21 @@ def get_data_dir() -> str:
 DATA_DIR = get_data_dir()
 DEFAULT_CACHE_CONNECTION = get_connection(MODULE_NAME)
 
+# Databases contained in PathMe
+#: KEGG
 KEGG = 'kegg'
 KEGG_DIR = os.path.join(DATA_DIR, KEGG)
 KEGG_BEL = os.path.join(KEGG_DIR, 'bel')
 KEGG_FILES = os.path.join(KEGG_DIR, 'xml')
 KEGG_CACHE = os.path.join(KEGG_DIR, 'cache')
 
+#: Reactome
 REACTOME = 'reactome'
 REACTOME_DIR = os.path.join(DATA_DIR, REACTOME)
 REACTOME_BEL = os.path.join(REACTOME_DIR, 'bel')
 REACTOME_FILES = os.path.join(REACTOME_DIR, 'rdf', 'biopax')
 
+#: WikiPathways
 WIKIPATHWAYS = 'wikipathways'
 WIKIPATHWAYS_DIR = os.path.join(DATA_DIR, WIKIPATHWAYS)
 WIKIPATHWAYS_BEL = os.path.join(WIKIPATHWAYS_DIR, 'bel')
@@ -68,26 +75,49 @@ def ensure_pathme_folders():
 UNKNOWN = 'unknown'
 
 # Other namespaces
+
+#: InterPro
 INTERPRO = 'interpro'
+#: Pfam
 PFAM = 'pfam'
+#: BRENDA
 BRENDA = 'brenda'
+#: ChEMBL
 CHEMBL = 'chembl'
+#: miRBase
 MIRBASE = 'mirbase'
 
-KEGG_ID = 'kegg_id'
-KEGG_NAME = 'kegg_name'
-KEGG_TYPE = 'kegg_type'
+#: HGNC
 HGNC = 'HGNC'
+#: HGNC Symbol
 HGNC_SYMBOL = 'HGNC symbol'
+#: ENSEMBL
 ENSEMBL = 'ENSEMBL'
+#: EXPASY
 EXPASY = 'EXPASY'
+#: ENTREZ
 ENTREZ = 'ENTREZ'
+#: UniProt
 UNIPROT = 'UniProt'
+#: ChEBI
 CHEBI = 'ChEBI'
+#: ChEBI chemical name
 CHEBI_NAME = 'ChEBI name'
+#: PubChem
 PUBCHEM = 'PubChem'
+#: Wikipedia
 WIKIPEDIA = 'WIKIPEDIA'
 
+#: KEGG entity identifier
+KEGG_ID = 'kegg_id'
+
+#: KEGG entity name
+KEGG_NAME = 'kegg_name'
+
+#: KEGG entity type
+KEGG_TYPE = 'kegg_type'
+
+#: Protein modification types present in KEGG
 KEGG_MODIFICATIONS = {
     'phosphorylation': 'Ph',
     'glycosylation': 'Glyco',
@@ -102,18 +132,21 @@ def ensure_pathme_folders():
     rna
 )
 
+#: KEGG reference Kanehisa, M., & Goto, S. (2000). KEGG: kyoto encyclopedia of genes and genomes. Nucleic acids research, 28(1), 27-30.
 KEGG_CITATION = '10592173'
+#: Reactome reference Fabregat, A., et al. (2018). The reactome pathway knowledgebase. Nucleic acids research, 46(D1), D649-D655.
 REACTOME_CITATION = '29145629'
 
-# FIXME why doesn't this just import the compath_resources package?
-KEGG_WIKIPATHWAYS_MAPPINGS = 'https://github.com/ComPath/curation/raw/master/mappings/kegg_wikipathways.xlsx'
-KEGG_REACTOME_MAPPINGS = 'https://github.com/ComPath/curation/raw/master/mappings/kegg_reactome.xlsx'
-WIKIPATHWAYS_REACTOME_MAPPINGS = 'https://github.com/ComPath/curation/raw/master/mappings/wikipathways_reactome.xlsx'
-
+#: REST API to KEGG
 KEGG_KGML_URL = 'http://rest.kegg.jp/get/{}/kgml'
+
+#: Reactome RDF
 RDF_REACTOME = 'ftp://ftp.ebi.ac.uk/pub/databases/RDF/reactome/r71/reactome-biopax.tar.bz2'
+
+#: WikiPathways RDF
 RDF_WIKIPATHWAYS = 'http://data.wikipathways.org/20190910/rdf/wikipathways-20190910-rdf-wp.zip'
 
+#: Mapping to compare conversion of entities from KEGG XML (i.e. KGML) to BEL
 KEGG_STATS_COLUMN_NAMES = {
     'nodes': 'BEL Nodes',
     'entities': 'XML Entities',
@@ -159,6 +192,7 @@ def ensure_pathme_folders():
     'brite': 'XML Brite Hierarchy',
 }
 
+#: Mapping of BEL terms to their definitions
 BEL_STATS_COLUMN_NAMES = {
     'nodes': 'BEL Nodes',
     'edges': 'BEL Edges',

diff --git a/src/pathme/kegg/kegg_xml_parser.py b/src/pathme/kegg/kegg_xml_parser.py
@@ -13,7 +13,7 @@
 from bio2bel_kegg.constants import API_KEGG_GET
 from bio2bel_kegg.parsers.description import parse_description
 
-from pathme.constants import CHEBI, CHEBI_NAME, HGNC, HGNC_SYMBOL, KEGG_CACHE, PUBCHEM, UNIPROT
+from pathme.constants import CHEBI, CHEBI_NAME, HGNC, HGNC_SYMBOL, KEGG_CACHE, PUBCHEM, UNIPROT, KEGG_ID, KEGG_TYPE
 from pathme.wikipathways.utils import merge_two_dicts
 
 log = logging.getLogger(__name__)
@@ -124,8 +124,8 @@ def _process_kegg_api_get_entity(entity, type, hgnc_manager, chebi_manager):
 
     node_dict = _post_process_api_query(node_meta_data, hgnc_manager, chebi_manager)
 
-    node_dict['kegg_id'] = entity
-    node_dict['kegg_type'] = type
+    node_dict[KEGG_ID] = entity
+    node_dict[KEGG_TYPE] = type
 
     with open(_entity_filepath, 'w') as f:
         json.dump(node_dict, f)
@@ -172,7 +172,7 @@ def get_entity_nodes(tree, hgnc_manager, chebi_manager):
 
         elif kegg_type.startswith('map'):
 
-            map_info = {'kegg_id': kegg_ids}
+            map_info = {KEGG_ID: kegg_ids}
 
             for graphics in entry.iter('graphics'):
                 map_name = graphics.get('name')
@@ -184,8 +184,8 @@ def get_entity_nodes(tree, hgnc_manager, chebi_manager):
 
             for ortholog_id in kegg_ids.split(' '):
                 ortholog_info = {
-                    'kegg_id': ortholog_id,
-                    'kegg_type': kegg_type
+                    KEGG_ID: ortholog_id,
+                    KEGG_TYPE: kegg_type
                 }
 
                 ortholog_dict[entry_id].append(ortholog_info)