Merge pull request #227 from BioAnalyticResource/dev

Update from dev branch

Author: asherpasha

api/resources/gene_information.py

@@ -8,14 +8,14 @@
 from api.models.eplant_soybean import Isoforms as EPlantSoybeanIsoforms
 from api.utils.bar_utils import BARUtils
 from marshmallow import Schema, ValidationError, fields as marshmallow_fields
-from api import cache, db
+from api import db
 
 
 gene_information = Namespace("Gene Information", description="Information about Genes", path="/gene_information")
 
 # I think this is only needed for Swagger UI POST
-gene_isoforms_request_fields = gene_information.model(
-    "GeneIsoforms",
+gene_information_request_fields = gene_information.model(
+    "GeneInformation",
     {
         "species": fields.String(required=True, example="arabidopsis"),
         "genes": fields.List(
@@ -28,46 +28,62 @@
 
 
 # Validation is done in a different way to keep things simple
-class GeneIsoformsSchema(Schema):
+class GeneInformationSchema(Schema):
     species = marshmallow_fields.String(required=True)
     genes = marshmallow_fields.List(cls_or_instance=marshmallow_fields.String)
 
 
-@gene_information.route("/gene_alias")
-class GeneAliasList(Resource):
-    def get(self):
-        """This end point returns the list of species available"""
-        species = ["arabidopsis"]  # This are the only species available so far
-        return BARUtils.success_exit(species)
-
+@gene_information.route("/gene_aliases")
+class GeneAliases(Resource):
+    @gene_information.expect(gene_information_request_fields)
+    def post(self):
+        """This end point retrieves gene aliases for a large dataset"""
+        json_data = request.get_json()
+        data = {}
 
-@gene_information.route("/gene_alias/<string:species>/<string:gene_id>")
-class GeneAlias(Resource):
-    @gene_information.param("species", _in="path", default="arabidopsis")
-    @gene_information.param("gene_id", _in="path", default="At3g24650")
-    @cache.cached()
-    def get(self, species="", gene_id=""):
-        """This end point provides gene alias given a gene ID."""
-        aliases = []
+        # Validate json
+        try:
+            json_data = GeneInformationSchema().load(json_data)
+        except ValidationError as err:
+            return BARUtils.error_exit(err.messages), 400
 
-        # Escape input
-        species = escape(species)
-        gene_id = escape(gene_id)
+        genes = json_data["genes"]
+        species = json_data["species"]
 
+        # Set species and check gene ID format
         if species == "arabidopsis":
-            if BARUtils.is_arabidopsis_gene_valid(gene_id):
-                rows = db.session.execute(db.select(AgiAlias).where(AgiAlias.agi == gene_id)).scalars().all()
-                [aliases.append(row.alias) for row in rows]
-            else:
-                return BARUtils.error_exit("Invalid gene id"), 400
+            database = AgiAlias
+
+            # Check if gene is valid
+            for gene in genes:
+                if not BARUtils.is_arabidopsis_gene_valid(gene):
+                    return BARUtils.error_exit("Invalid gene id"), 400
+
         else:
-            return BARUtils.error_exit("No data for the given species")
+            return BARUtils.error_exit("Invalid species"), 400
+
+        # Query must be run individually for each species
+        rows = db.session.execute(db.select(database).where(database.agi.in_(genes))).scalars().all()
+
+        # If there are any isoforms found, return data
+        data = []
+        data_items = {}
+
+        if len(rows) > 0:
+            for row in rows:
+                if row.agi in data_items.keys():
+                    data_items[row.agi].append(row.agi)
+                else:
+                    data_items[row.agi] = []
+                    data_items[row.agi].append(row.alias)
+
+            for gene in data_items.keys():
+                data.append({"gene": gene, "aliases": data_items[gene]})
+
+            return BARUtils.success_exit(data)
 
-        # Return results if there are data
-        if len(aliases) > 0:
-            return BARUtils.success_exit(aliases)
         else:
-            return BARUtils.error_exit("There are no data found for the given gene")
+            return BARUtils.error_exit("No data for the given species/genes"), 400
 
 
 @gene_information.route("/gene_isoforms/<string:species>/<string:gene_id>")
@@ -126,7 +142,7 @@ def get(self, species="", gene_id=""):
 
 @gene_information.route("/gene_isoforms/")
 class PostGeneIsoforms(Resource):
-    @gene_information.expect(gene_isoforms_request_fields)
+    @gene_information.expect(gene_information_request_fields)
     def post(self):
         """This end point returns gene isoforms data for a multiple genes for a species.
         Only genes/isoforms with pdb structures are returned"""
@@ -136,7 +152,7 @@ def post(self):
 
         # Validate json
         try:
-            json_data = GeneIsoformsSchema().load(json_data)
+            json_data = GeneInformationSchema().load(json_data)
         except ValidationError as err:
             return BARUtils.error_exit(err.messages), 400
 

config/databases/annotations_lookup.sql

@@ -1,13 +1,13 @@
--- MariaDB dump 10.17  Distrib 10.4.12-MariaDB, for OpenBSD (amd64)
+-- MySQL dump 10.13  Distrib 8.3.0, for Linux (x86_64)
 --
 -- Host: localhost    Database: annotations_lookup
 -- ------------------------------------------------------
--- Server version	10.4.12-MariaDB
+-- Server version	8.3.0
 
 /*!40101 SET @OLD_CHARACTER_SET_CLIENT=@@CHARACTER_SET_CLIENT */;
 /*!40101 SET @OLD_CHARACTER_SET_RESULTS=@@CHARACTER_SET_RESULTS */;
 /*!40101 SET @OLD_COLLATION_CONNECTION=@@COLLATION_CONNECTION */;
-/*!40101 SET NAMES utf8mb4 */;
+/*!50503 SET NAMES utf8mb4 */;
 /*!40103 SET @OLD_TIME_ZONE=@@TIME_ZONE */;
 /*!40103 SET TIME_ZONE='+00:00' */;
 /*!40014 SET @OLD_UNIQUE_CHECKS=@@UNIQUE_CHECKS, UNIQUE_CHECKS=0 */;
@@ -19,7 +19,7 @@
 -- Current Database: `annotations_lookup`
 --
 
-CREATE DATABASE /*!32312 IF NOT EXISTS*/ `annotations_lookup` /*!40100 DEFAULT CHARACTER SET latin1 */;
+CREATE DATABASE /*!32312 IF NOT EXISTS*/ `annotations_lookup` /*!40100 DEFAULT CHARACTER SET latin1 */ /*!80016 DEFAULT ENCRYPTION='N' */;
 
 USE `annotations_lookup`;
 
@@ -29,7 +29,7 @@ USE `annotations_lookup`;
 
 DROP TABLE IF EXISTS `agi_alias`;
 /*!40101 SET @saved_cs_client     = @@character_set_client */;
-/*!40101 SET character_set_client = utf8 */;
+/*!50503 SET character_set_client = utf8mb4 */;
 CREATE TABLE `agi_alias` (
   `agi` varchar(30) NOT NULL,
   `alias` varchar(30) NOT NULL,
@@ -45,7 +45,7 @@ CREATE TABLE `agi_alias` (
 
 LOCK TABLES `agi_alias` WRITE;
 /*!40000 ALTER TABLE `agi_alias` DISABLE KEYS */;
-INSERT INTO `agi_alias` VALUES ('At3g24650','ABI3','2019-04-02'),('At3g24650','AtABI3','2019-04-02'),('At3g24650','SIS10','2019-04-02');
+INSERT INTO `agi_alias` VALUES ('At1g01010','ANAC001','2022-06-30'),('At1g01010','NAC001','2022-06-30'),('At1g01010','NTL10','2022-06-30'),('At1g01020','ARV1','2022-06-30');
 /*!40000 ALTER TABLE `agi_alias` ENABLE KEYS */;
 UNLOCK TABLES;
 /*!40103 SET TIME_ZONE=@OLD_TIME_ZONE */;
@@ -58,4 +58,4 @@ UNLOCK TABLES;
 /*!40101 SET COLLATION_CONNECTION=@OLD_COLLATION_CONNECTION */;
 /*!40111 SET SQL_NOTES=@OLD_SQL_NOTES */;
 
--- Dump completed on 2020-06-26 15:59:53
+-- Dump completed on 2024-05-02 19:19:37

docs/requirements.txt

@@ -4,7 +4,7 @@ beautifulsoup4==4.12.3
 certifi==2024.2.2
 charset-normalizer==3.3.2
 docutils==0.21.2
-furo==2024.1.29
+furo==2024.4.27
 idna==3.7
 imagesize==1.4.1
 Jinja2==3.1.3

docs/source/user_guide.rst

@@ -1,4 +1,29 @@
 User Guide
 ==========
 
-Coming soon ...
+Gene Information
+----------------
+
+These endpoints can be used to retrieve gene annotations, aliases, probesets etc. For large input datasets, please use
+POST request because POST request do not have URL length limit. Currently, the following species are available.
+
+1. Arabidopsis
+
+**POST /gene_information/gene_aliases/**
+
+This end point provides gene aliases given a list of gene IDs for a species.
+
+**GET /gene_information/gene_isoforms/{species}/{gene_id}**
+
+This end point provides gene isoforms given a gene ID.
+
+**POST /gene_information/gene_isoforms/**
+
+This end point provides gene isoforms given a gene ID. This is the POST request that take a JSON object of species and
+genes.
+
+
+
+
+
+

requirements.txt

@@ -1,19 +1,19 @@
 aniso8601==9.0.1
 async-timeout==4.0.3
 attrs==23.2.0
-black==24.4.1
-blinker==1.7.0
+black==24.4.2
+blinker==1.8.2
 cachelib==0.9.0
-certifi==2024.2.2
+certifi==2024.7.4
 charset-normalizer==3.3.2
 click==8.1.7
-coverage==7.5.0
+coverage==7.5.4
 Deprecated==1.2.14
-flake8==7.0.0
+flake8==7.1.0
 Flask==3.0.3
-Flask-Caching==2.1.0
-Flask-Cors==4.0.0
-Flask-Limiter==3.6.0
+Flask-Caching==2.3.0
+Flask-Cors==4.0.1
+Flask-Limiter==3.7.0
 flask-marshmallow==1.2.1
 flask-restx==1.3.0
 Flask-SQLAlchemy==3.1.1
@@ -22,37 +22,39 @@ idna==3.7
 importlib_resources==6.4.0
 iniconfig==2.0.0
 itsdangerous==2.2.0
-Jinja2==3.1.3
-jsonschema==4.21.1
+Jinja2==3.1.4
+jsonschema==4.23.0
 jsonschema-specifications==2023.12.1
-limits==3.11.0
+limits==3.13.0
 markdown-it-py==3.0.0
 MarkupSafe==2.1.5
-marshmallow==3.21.1
+marshmallow==3.21.3
 mccabe==0.7.0
 mdurl==0.1.2
 mypy-extensions==1.0.0
 mysqlclient==2.2.4
 ordered-set==4.1.0
-packaging==24.0
+packaging==24.1
 pathspec==0.12.1
-platformdirs==4.2.1
+platformdirs==4.2.2
 pluggy==1.5.0
-pycodestyle==2.11.1
+pycodestyle==2.12.0
 pyflakes==3.2.0
-Pygments==2.17.2
+Pygments==2.18.0
 pyrsistent==0.20.0
-pytest==8.1.1
+pytest==8.2.2
 python-dateutil==2.9.0.post0
 pytz==2024.1
-redis==5.0.4
-referencing==0.35.0
-requests==2.31.0
+redis==5.0.7
+referencing==0.35.1
+requests==2.32.3
 rich==13.7.1
-rpds-py==0.18.0
+rpds-py==0.19.0
+setuptools==70.2.0
 six==1.16.0
-SQLAlchemy==2.0.29
-typing_extensions==4.11.0
-urllib3==2.2.1
-Werkzeug==3.0.2
+SQLAlchemy==2.0.31
+typing_extensions==4.12.2
+urllib3==2.2.2
+Werkzeug==3.0.3
+wheel==0.43.0
 wrapt==1.16.0

tests/resources/test_gene_information.py

@@ -6,39 +6,41 @@ class TestIntegrations(TestCase):
     def setUp(self):
         self.app_client = app.test_client()
 
-    def test_get_gene_alias_list(self):
-        """This function tests the gene alias list get function
+    def test_post_gene_aliases(self):
+        """This test POST request to /gene_information/gene_aliases/
         :return:
         """
-        response = self.app_client.get("/gene_information/gene_alias")
-        expected = {"wasSuccessful": True, "data": ["arabidopsis"]}
-        self.assertEqual(response.json, expected)
-
-    def test_get_arabidopsis_gene_alias(self):
-        """This tests checks GET request for gene alias Arabidopsis
-        :return:
-        """
-        # Valid data
-        response = self.app_client.get("/gene_information/gene_alias/arabidopsis/At3g24650")
-        expected = {"wasSuccessful": True, "data": ["ABI3", "AtABI3", "SIS10"]}
-        self.assertEqual(response.json, expected)
-
-        # Data not found, but gene is valid
-        response = self.app_client.get("/gene_information/gene_alias/arabidopsis/At3g24651")
+        data = {"species": "arabidopsis", "genes": ["AT1G01010", "AT1G01020"]}
         expected = {
-            "wasSuccessful": False,
-            "error": "There are no data found for the given gene",
+            "wasSuccessful": True,
+            "data": [
+                {"gene": "At1g01010", "aliases": ["ANAC001", "At1g01010", "At1g01010"]},
+                {"gene": "At1g01020", "aliases": ["ARV1"]},
+            ],
         }
+        response = self.app_client.post("/gene_information/gene_aliases", json=data)
+        self.assertEqual(response.status_code, 200)
         self.assertEqual(response.json, expected)
 
-        # Invalid Gene
-        response = self.app_client.get("/gene_information/gene_alias/arabidopsis/At3g2465x")
+        # Invalid gene
+        data = {"species": "abc", "genes": ["AT1G01010", "AT1G01020"]}
+        expected = {"wasSuccessful": False, "error": "Invalid species"}
+        response = self.app_client.post("/gene_information/gene_aliases", json=data)
+        self.assertEqual(response.status_code, 400)
+        self.assertEqual(response.json, expected)
+
+        # Invalid gene
+        data = {"species": "arabidopsis", "genes": ["abc", "AT1G01020"]}
         expected = {"wasSuccessful": False, "error": "Invalid gene id"}
+        response = self.app_client.post("/gene_information/gene_aliases", json=data)
+        self.assertEqual(response.status_code, 400)
         self.assertEqual(response.json, expected)
 
-        # Invalid Species
-        response = self.app_client.get("/gene_information/gene_alias/x/At3g24650")
-        expected = {"wasSuccessful": False, "error": "No data for the given species"}
+        # No data
+        data = {"species": "arabidopsis", "genes": ["AT1G01011", "AT1G01021"]}
+        expected = {"wasSuccessful": False, "error": "No data for the given species/genes"}
+        response = self.app_client.post("/gene_information/gene_aliases", json=data)
+        self.assertEqual(response.status_code, 400)
         self.assertEqual(response.json, expected)
 
     def test_get_arabidopsis_gene_isoform(self):