Merge pull request #7 from smart-on-fhir/mg/subtables

Added convenience subtables

.github/workflows/ci.yaml

@@ -9,7 +9,9 @@ jobs:
       - name: Set up Python 
         uses: actions/setup-python@v5
         with:
-          python-version: '3.10'
+          python-version: '3.11'
+      - name: Get library from main
+        run: pip install git+https://github.com/smart-on-fhir/cumulus-library.git
       - name: Install linters
         run: |
           python -m pip install --upgrade pip
@@ -29,7 +31,10 @@ jobs:
       - name: Set up Python 
         uses: actions/setup-python@v5
         with:
-          python-version: '3.10'
+          python-version: '3.11'
+
+      - name: Get library from main
+        run: pip install git+https://github.com/smart-on-fhir/cumulus-library.git
 
       - name: Install dependencies
         run: |

README.md

@@ -23,6 +23,17 @@ Note: This study is explicitly namespaced in its own schema, `umls`. Make sure y
 database is not using this schema for another use. Do not create tables inside this
 schema by another means.
 
+## Additional custom tables
+
+The following tables are a derived from the primary tables, and are included here as a 
+convenience to avoid having to compute these on a repeated basis
+
+- **mrrel_is_a** a subset of the relationships in mrrel, including only those that define
+that concept A is a member of concept B (i.e. is a child, or is explicitly marked as
+being a tradename/member belonging to the parent concept).
+- **mrconso_drugs** a subset of the entity list in mrconso, limited to vocabularies
+specifically dealing with drug identifiers (i.e. SNOMED, RxNorm, etc.)
+
 ## Licensing details
 
 The `cumulus-library-umls` study is provided as a convenience to install the

cumulus_library_umls/__init__.py

@@ -0,0 +1 @@
+__version__ = "1.0.0"

cumulus_library_umls/umls/umls_builder.py

@@ -1,7 +1,7 @@
 import pathlib
 
 import pandas
-from cumulus_library import base_table_builder, base_utils, databases
+from cumulus_library import base_table_builder, base_utils, study_manifest
 from cumulus_library.apis import umls
 from cumulus_library.template_sql import base_templates
 
@@ -145,10 +145,9 @@ def create_parquet(
 
     def prepare_queries(
         self,
-        cursor: databases.DatabaseCursor,
-        schema: str,
+        config: base_utils.StudyConfig,
+        manifest: study_manifest.StudyManifest,
         *args,
-        config=base_utils.StudyConfig,
         **kwargs,
     ):
         download_path = pathlib.Path(__file__).resolve().parent / "downloads"
@@ -184,7 +183,7 @@ def prepare_queries(
                     )
                     self.queries.append(
                         base_templates.get_ctas_from_parquet_query(
-                            schema_name=schema,
+                            schema_name=config.schema,
                             table_name=f"umls__{file.stem}",
                             local_location=parquet_path / f"{file.stem}.parquet",
                             remote_location=remote_path,
@@ -193,3 +192,17 @@ def prepare_queries(
                         )
                     )
                     progress.advance(task)
+
+        # Section for resuable cross-study helper tables
+        self.queries.append(
+            """CREATE TABLE umls__mrrel_is_a AS
+SELECT * FROM umls.mrrel
+WHERE REL = 'CHD' 
+OR RELA in ('isa','tradename_of','has_tradename','has_basis_of_strength_substance')"""
+        )
+        self.queries.append(
+            """CREATE TABLE umls__mrconso_drugs AS
+SELECT * FROM umls.mrconso
+WHERE SAB in ('ATC','CVX','DRUGBANK','GS','MMSL','MMX','MTHCMSFRFMTHSPL','NDDF',
+    'RXNORM','SNOMEDCT_US','USP','VANDF')"""
+        )

pyproject.toml

@@ -1,9 +1,8 @@
 [project]
-version = "0.2.0"
 name = "cumulus-library-umls"
-requires-python = ">= 3.10"
+requires-python = ">= 3.11"
 dependencies = [
-    "cumulus-library >= 2.3",
+    "cumulus-library >= 3.0",
 ]
 description = "A Unified Medical Language System® Metathesaurus study for the Cumulus project"
 readme = "README.md"
@@ -15,6 +14,7 @@ classifiers = [
     "Programming Language :: Python :: 3",
     "Topic :: Software Development :: Libraries :: Python Modules",
 ]
+dynamic=["version"]
 [project.optional-dependencies]
 dev = [
     "ruff == 0.2.1",

tests/test_umls_builder.py

@@ -3,7 +3,7 @@
 
 import pytest
 import responses
-from cumulus_library import base_utils, databases, db_config
+from cumulus_library import base_utils, databases, db_config, study_manifest
 
 from cumulus_library_umls.umls import umls_builder
 
@@ -59,10 +59,13 @@ def test_create_query(mock_resolve, mock_responses, tmp_path):
 
     db_config.db_type = "duckdb"
     config = base_utils.StudyConfig(
-        db=databases.DuckDatabaseBackend(f"{tmp_path}/duckdb"), umls_key="123"
+        db=databases.DuckDatabaseBackend(f"{tmp_path}/duckdb"),
+        umls_key="123",
+        schema="main",
     )
     builder = umls_builder.UMLSBuilder()
-    builder.prepare_queries(cursor=config.db.cursor(), schema="main", config=config)
+    manifest = study_manifest.StudyManifest()
+    builder.prepare_queries(config=config, manifest=manifest)
     expected = f"""CREATE TABLE IF NOT EXISTS umls__TESTTABLE AS SELECT
     TTY,
     CODE
@@ -88,10 +91,13 @@ def test_create_query_download_exists(mock_resolve, mock_responses, tmp_path):
 
     db_config.db_type = "duckdb"
     config = base_utils.StudyConfig(
-        db=databases.DuckDatabaseBackend(f"{tmp_path}/duckdb"), umls_key="123"
+        db=databases.DuckDatabaseBackend(f"{tmp_path}/duckdb"),
+        umls_key="123",
+        schema="main",
     )
     builder = umls_builder.UMLSBuilder()
-    builder.prepare_queries(cursor=config.db.cursor(), schema="main", config=config)
+    manifest = study_manifest.StudyManifest()
+    builder.prepare_queries(config=config, manifest=manifest)
     download_dirs = sorted((tmp_path / "downloads").iterdir())
     assert len(download_dirs) == 1
     assert "2000AA" in str(download_dirs[0])