Merge pull request #560 from Steinbeck-Lab/development

feat: SMARTS convertion

Author: Kohulan

.github/workflows/dev-build.yml

@@ -1,4 +1,3 @@
-
 # This worklflow will perform following actions when the code is pushed to development branch:
 # - Test linting with pylint.
 # - Fetch Latest release.
@@ -43,6 +42,16 @@ jobs:
           push: true
           build-args: |
             RELEASE_VERSION=dev-latest
-          tags: ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:dev-latest
-          username: ${{ env.DOCKER_HUB_USERNAME  }}
-          password: ${{ env.DOCKER_HUB_PASSWORD  }}
+          tags: ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:api-dev
+
+      - name: Build and push frontend Docker image
+        uses: docker/[email protected]
+        with:
+          context: ./frontend
+          file: ./frontend/Dockerfile
+          push: true
+          # REACT_APP_API_URL will use the default from frontend/Dockerfile (https://dev.api.naturalproducts.net/latest)
+          # If a specific one is needed for this dev build, add build-args here.
+          # build-args: |
+          #   REACT_APP_API_URL=your-dev-frontend-api-url
+          tags: ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:app-dev

.github/workflows/prod-build.yml

@@ -1,4 +1,3 @@
-
 # This worklflow will perform following actions when a release is published:
 # - Fetch Latest release.
 # - Build the latest docker image in production.
@@ -57,8 +56,6 @@ jobs:
           tags: |
             ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:${{ steps.fetch-latest-release.outputs.tag_name }}
             ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:latest
-          username: ${{ env.DOCKER_HUB_USERNAME }}
-          password: ${{ env.DOCKER_HUB_PASSWORD }}
 
       - name: Build and push lite Docker image
         uses: docker/build-push-action@v4
@@ -71,5 +68,15 @@ jobs:
           tags: |
             ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:${{ steps.fetch-latest-release.outputs.tag_name }}-lite
             ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:latest-lite
-          username: ${{ env.DOCKER_HUB_USERNAME }}
-          password: ${{ env.DOCKER_HUB_PASSWORD }}
+
+      - name: Build and push frontend Docker image
+        uses: docker/build-push-action@v4
+        with:
+          context: ./frontend
+          file: ./frontend/Dockerfile
+          push: true
+          build-args: |
+            REACT_APP_API_URL=/api/latest # For prod Traefik setup
+          tags: |
+            ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:app-${{ steps.fetch-latest-release.outputs.tag_name }}
+            ${{ env.REPOSITORY_NAMESPACE }}/${{ env.REPOSITORY_NAME }}:app

.github/workflows/test.yml

@@ -31,8 +31,8 @@ jobs:
         python -m pip install --upgrade pip
         pip3 install --upgrade setuptools pip
         pip3 install --no-cache-dir -r requirements.txt
-        pip3 install --no-deps decimer-segmentation==1.1.3
-        pip3 install --no-deps decimer==2.3.0
+        pip install git+https://github.com/Kohulan/DECIMER-Image-Segmentation.git@bbox --no-deps
+        pip3 install --no-deps decimer
         pip3 install --no-deps STOUT-pypi==2.0.5
         pip install flake8 pytest
         pip install pytest-cov

Dockerfile

@@ -2,7 +2,7 @@ FROM continuumio/miniconda3:24.1.2-0 AS cheminf-python-ms
 
 ENV PYTHON_VERSION=3.11 \
     INCLUDE_OCSR=true \
-    JAVA_HOME=/usr/lib/jvm/java-11-openjdk-amd64/ \
+    JAVA_HOME=/usr/lib/jvm/java-11-openjdk-arm64/ \
     # Add default number of workers
     WORKERS=2 \
     # Add other Python configurations

Dockerfile.lite

@@ -2,7 +2,7 @@ FROM continuumio/miniconda3:24.1.2-0 AS cheminf-python-ms
 
 ENV PYTHON_VERSION=3.11 \
     INCLUDE_OCSR=false \
-    JAVA_HOME=/usr/lib/jvm/java-11-openjdk-amd64/ \
+    JAVA_HOME=/usr/lib/jvm/java-11-openjdk-arm64/ \
     # Add default number of workers
     WORKERS=1 \
     # Add other Python configurations

app/routers/chem.py

@@ -977,6 +977,11 @@ async def all_filter_molecules(
         title="NPlikenessScore",
         description="Calculate NPlikenessScore in the range (e.g., 0-10)",
     ),
+    filterOperator: Literal["AND", "OR"] = Query(
+        "OR",
+        title="Filter Operator",
+        description="Logic for combining filter results: AND requires all selected filters to pass, OR requires at least one filter to pass",
+    ),
 ):
     all_smiles = []
     for item in io.StringIO(smiles_list):
@@ -1051,12 +1056,30 @@ async def all_filter_molecules(
                 else:
                     results.append("False")
 
-            output_list = [
-                x if x != "True" and x != "False" else ("T" if x == "True" else "F")
-                for x in results
-            ]
-            final_results = ", ".join(output_list).replace(":,", " : ")
-            all_smiles.append(final_results)
+            # Filter results based on AND/OR operator
+            # First element is the SMILES, so we start checking from index 1
+            filter_results = (
+                [r == "True" for r in results[1:]] if len(results) > 1 else []
+            )
+
+            # Apply filter operator logic
+            passes_filters = False
+            if filter_results:
+                if filterOperator == "AND":
+                    passes_filters = all(filter_results)  # All filters must pass
+                else:  # OR
+                    passes_filters = any(
+                        filter_results
+                    )  # At least one filter must pass
+
+            # Only include SMILES that pass the filter operator logic
+            if not filter_results or passes_filters:
+                output_list = [
+                    x if x != "True" and x != "False" else ("T" if x == "True" else "F")
+                    for x in results
+                ]
+                final_results = ", ".join(output_list).replace(":,", " : ")
+                all_smiles.append(final_results)
 
     return all_smiles
 

app/routers/converters.py

@@ -37,6 +37,7 @@
 from app.schemas.converters_schema import GenerateSMILESResponse
 from app.schemas.converters_schema import ThreeDCoordinatesResponse
 from app.schemas.converters_schema import TwoDCoordinatesResponse
+from app.schemas.converters_schema import GenerateSMARTSResponse
 from app.schemas.error import BadRequestModel
 from app.schemas.error import ErrorResponse
 from app.schemas.error import NotFoundModel
@@ -702,3 +703,45 @@ async def smiles_convert_to_formats(
             status_code=422,
             detail="Error processing request: " + str(e),
         )
+
+
+@router.get(
+    "/smarts",
+    summary="Generate SMARTS from a given SMILES",
+    responses={
+        200: {
+            "description": "Successful response",
+            "model": GenerateSMARTSResponse,
+        },
+        400: {"description": "Bad Request", "model": BadRequestModel},
+        404: {"description": "Not Found", "model": NotFoundModel},
+        422: {"description": "Unprocessable Entity", "model": ErrorResponse},
+    },
+)
+async def smiles_to_smarts(
+    smiles: str = Query(
+        title="SMILES",
+        description="SMILES representation of the molecule",
+        openapi_examples={
+            "example1": {
+                "summary": "Example: Caffeine",
+                "value": "CN1C=NC2=C1C(=O)N(C(=O)N2C)C",
+            },
+            "example2": {
+                "summary": "Example: Topiramate-13C6",
+                "value": "CC1(C)OC2COC3(COS(N)(=O)=O)OC(C)(C)OC3C2O1",
+            },
+        },
+    ),
+    toolkit: Literal["rdkit"] = Query(
+        default="rdkit",
+        description="Cheminformatics toolkit used in the backend",
+    ),
+):
+
+    if toolkit == "rdkit":
+        mol = parse_input(smiles, "rdkit", False)
+        if mol:
+            smarts = Chem.MolToSmarts(mol)
+            if smarts:
+                return str(smarts)

app/schemas/converters_schema.py

@@ -90,7 +90,7 @@ class GenerateSMILESResponse(BaseModel):
 
     smiles: str = Field(
         ...,
-        title="SMILES",
+        title="SMARTS",
         description="The generated SMILES string corresponding to the input text.",
     )
 
@@ -271,10 +271,10 @@ class GenerateSELFIESResponse(BaseModel):
     """Represents a response containing a generated SELFIES string.
 
     Properties:
-    - iupac (str): The generated SELFIES string.
+    - selfies (str): The generated SELFIES string.
     """
 
-    iupac: str = Field(
+    selfies: str = Field(
         ...,
         title="SMILES",
         description="The generated SELFIES string corresponding to the input SMILES.",
@@ -298,6 +298,37 @@ class Config:
         }
 
 
+class GenerateSMARTSResponse(BaseModel):
+    """Represents a response containing a generated SMARTS string.
+
+    Properties:
+    - smarts (str): The generated SMARTS string.
+    """
+
+    smarts: str = Field(
+        ...,
+        title="SMILES",
+        description="The generated SMARTS string corresponding to the input SMILES.",
+    )
+
+    class Config:
+        """Pydantic model configuration.
+
+        JSON Schema Extra:
+        - Includes examples of the response structure.
+        """
+
+        json_schema_extra = {
+            "examples": [
+                {
+                    "input": "CN1C(=O)C2=C(N=CN2C)N(C)C1=O",
+                    "message": "Success",
+                    "output": "[#6]-[#7]1:[#6]:[#7]:[#6]2:[#6]:1:[#6](=[#8]):[#7](:[#6](=[#8]):[#7]:2-[#6])-[#6]",
+                },
+            ],
+        }
+
+
 class GenerateFormatsResponse(BaseModel):
     """Represents a response containing a generated SELFIES string.
 

docker-compose.lite.yml

@@ -0,0 +1,67 @@
+services:
+  api:
+    build:
+      context: ./
+      dockerfile: Dockerfile.lite
+    container_name: cheminformatics-microservice-lite
+    environment:
+      - HOMEPAGE_URL=https://docs.api.naturalproducts.net
+      - RELEASE_VERSION=v2.6.0
+      - WORKERS=2  # Make workers configurable
+      - INCLUDE_OCSR=false
+    ports:
+      - "80:80"
+    healthcheck:
+      test: curl -f http://localhost:80/latest/chem/health || exit 1
+      interval: 90s
+      timeout: 10s
+      retries: 20
+      start_period: 60s
+    restart: unless-stopped
+    security_opt:
+      - no-new-privileges:true
+    networks:
+      - cm_fastapi
+
+  prometheus:
+    image: prom/prometheus:latest  # Specify version for better stability
+    container_name: prometheus
+    ports:
+      - "9090:9090"
+    volumes:
+      - ./prometheus_data/prometheus.yml:/etc/prometheus/prometheus.yml:ro
+      - prometheus_data:/prometheus
+    command:
+      - '--config.file=/etc/prometheus/prometheus.yml'
+      - '--storage.tsdb.path=/prometheus'
+      - '--web.console.libraries=/usr/share/prometheus/console_libraries'
+      - '--web.console.templates=/usr/share/prometheus/consoles'
+    restart: unless-stopped
+    networks:
+      - cm_fastapi
+
+  grafana:
+    image: grafana/grafana:latest
+    container_name: grafana
+    user: "472"  # Grafana's official user ID
+    ports:
+      - "3000:3000"
+    volumes:
+      - grafana_data:/var/lib/grafana
+    environment:
+      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_ADMIN_PASSWORD:-admin}  # Configurable password
+      - GF_USERS_ALLOW_SIGN_UP=false
+    restart: unless-stopped
+    depends_on:
+      - prometheus
+    networks:
+      - cm_fastapi
+
+volumes:
+  prometheus_data:
+  grafana_data:
+
+networks:
+  cm_fastapi:
+    name: cm_fastapi
+    driver: bridge