PES Adjustments, Check is User is in enviPath Entra Group

Initial bayer app

Show Pack Classification

Adjusted docker compose to bayer specifics

Adjusted Dockerfile for Bayer

Adding secret flags to group, add secret pools to packages

Adjusted View for Package creation

Prep configs, added Package Create Modal

wip

More on PES

wip

wip

Wip

minor

PW interactions

API PES

wip

Make Select Widget reflect required

make required generallay available

Update UI if pathway mode is set to build

Added ais

circle adjustments

Initial Zoom, fix AD Creation

wip

auth log, bb4g fix

missing import

Added viz hint if PES is part of reaction

Add Edge check for pes

flip boolean

...

pes

Added extra

...

In / Out Edges Viz, Submitting Button Text

...

Make PES Link clickable

Return proper http response instead of error

Fixed error return, removed unused options

Fix PES Link HTML for other entities

Fixed molfile assignment, adjusted Export

Package Export/Import cycle

highlight Description links

implemented non persistent

Harmonised proposed field in Json output

Added pesLink field to PW Api output

PES Fields in API Output

removed debug

Fix Classification import, Fix PES Deserialization

underline pes link in templates

Fix alter name/desc for node, make /node /edge funcitonal

provide setting link and copy button

Implemented Compound Names / Reaction Names View Option

Unconnected Nodes

Make links thicker, reduce timeout trigger time

Show proposed info in popover

Pathway Build no stereo removal

Include probs in reaction name option viz

Detect clicks outside nodes/edges

Provide proper Error Pages

View Package Perm

wip

sync

Auth log

auth log leftovers

...

secret packs viz

auth log for api

model stats

Fix Package Adjustment

Adjust Group Auth Log

Fix Secret image size in Navbar

leftover

...

minor

.gitea/workflows/build-image.yaml

@@ -0,0 +1,62 @@
+name: Build Docker Image
+
+# Trigger when a PR to main/develop is completed.
+on:
+  pull_request:
+    branches:
+      - main
+      - develop
+    types:
+      - closed
+
+jobs:
+  build-and-push:
+    if: ${{ github.event.pull_request.merged == true }}
+    runs-on: ubuntu-latest
+    steps:
+      # Fetch the repository content for the Docker build context.
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      # Enable Buildx for BuildKit features (incl. SSH mount support).
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      # Authenticate against the container registry before pushing images.
+      - name: Log in to container registry
+        uses: docker/login-action@v3
+        with:
+          registry: git.envipath.com
+          username: ${{ secrets.CI_REGISTRY_USER }}
+          password: ${{ secrets.CI_REGISTRY_PASSWORD }}
+
+      # Generate image tags/labels:
+      # - PRs targeting main get "latest" and "main-sha"
+      # - PRs targeting develop get "dev" and "dev-sha"
+      - name: Extract metadata
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: git.envipath.com/envipath/envipy
+          tags: |
+            type=raw,value=latest,enable=${{ github.event.pull_request.base.ref == 'main' }}
+            type=sha,prefix=main-,enable=${{ github.event.pull_request.base.ref == 'main' }}
+            type=raw,value=dev,enable=${{ github.event.pull_request.base.ref == 'develop' }}
+            type=sha,prefix=dev-,enable=${{ github.event.pull_request.base.ref == 'develop' }}
+
+      # Load SSH key so Docker can pull private git+ssh dependencies during build.
+      - name: Setup SSH for private git dependencies
+        uses: webfactory/ssh-agent@v0.9.0
+        with:
+          ssh-private-key: ${{ secrets.ENVIPY_CI_PRIVATE_KEY }}
+
+      # Build and push the production image; forward SSH agent without registry cache reuse.
+      - name: Build and push Docker image
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          file: Dockerfile
+          push: true
+          ssh: default
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}

Dockerfile

@@ -60,6 +60,10 @@ COPY tests tests
 COPY utilities utilities
 COPY manage.py .
 
+# Used to run migrations etc
+COPY entrypoint.sh entrypoint.sh
+RUN chmod +x entrypoint.sh
+
 # Install frontend deps
 COPY package.json pnpm-lock.yaml pnpm-workspace.yaml ./
 
@@ -81,6 +85,7 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
     libxext6 \
     libfontconfig1 \
     nano \
+    openjdk-21-jre-headless \
     && rm -rf /var/lib/apt/lists/*
 
 RUN useradd -ms /bin/bash django
@@ -102,4 +107,5 @@ USER django
 
 EXPOSE 8000
 
-CMD ["gunicorn", "envipath.wsgi:application", "--bind", "0.0.0.0:8000", "--workers", "3"]
+ENTRYPOINT ["/app/entrypoint.sh"]
+CMD ["gunicorn", "envipath.wsgi:application", "--bind", "0.0.0.0:8000", "--workers", "8"]

bayer/templates/modals/collections/new_package_modal.html

@@ -5,11 +5,11 @@
   class="modal"
   x-data="{
     isSubmitting: false,
-    packageClassification: null,
+    packageClassification: '',
 
     reset() {
       this.isSubmitting = false;
-      this.packageClassification = null;
+      this.packageClassification = '';
     },
 
     setFormData(data) {
@@ -114,7 +114,7 @@
             x-model="packageClassification"
             required
           >
-            <option value="null" disabled selected>Select Classification</option>
+            <option value="" disabled>Select Classification</option>
             <option value="0">Internal</option>
             <option value="10">Restricted</option>
             <option value="20">Secret</option>

bayer/views.py

@@ -8,7 +8,7 @@ from django.shortcuts import redirect
 
 from bayer.models import PESCompound
 from epdb.logic import PackageManager
-from epdb.models import Pathway, Node
+from epdb.models import Pathway, Node, Group
 from epdb.views import _anonymous_or_real, error
 from utilities.decorators import package_permission_required
 
@@ -18,6 +18,23 @@ Package = s.GET_PACKAGE_MODEL()
 logger = logging.getLogger(__name__)
 
 
+def has_secret_group(user):
+    """
+    Determines if the specified user belongs to any secret group.
+
+    This function checks whether the given user is a member of any group
+    that is marked as secret.
+
+    Args:
+        user: The user for whom the check is performed.
+
+    Returns:
+        bool: True if the user belongs to at least one secret group,
+        False otherwise.
+    """
+    return Group.objects.filter(secret=True, user_member=user).exists()
+
+
 @package_permission_required()
 def create_pes(request, package_uuid):
     current_user = _anonymous_or_real(request)
@@ -38,7 +55,7 @@ def create_pes(request, package_uuid):
 
         if pes_link:
             try:
-                pes_data = fetch_pes(request, pes_link)
+                pes_data = fetch_pes(request, pes_link, current_user)
             except ValueError as e:
                 return error(
                     request,
@@ -98,7 +115,7 @@ def create_pes_node(request, package_uuid, pathway_uuid):
 
         if pes_link:
             try:
-                pes_data = fetch_pes(request, pes_link)
+                pes_data = fetch_pes(request, pes_link, current_user)
             except ValueError as e:
                 return error(
                     request,
@@ -157,16 +174,33 @@ def create_pes_node(request, package_uuid, pathway_uuid):
         return HttpResponseNotAllowed(["POST"])
 
 
-def fetch_pes(request, pes_url) -> dict:
+def get_application_token(prod: bool) -> str:
-    from epauth.views import get_access_token_from_request
+    scope = f"{s.PROD_PES_SCOPE if prod else s.NON_PROD_PES_SCOPE}/.default"
-    token = get_access_token_from_request(request)
 
-    if token is None:
+    url = f"https://login.microsoftonline.com/{s.MS_TENANT_ID}/oauth2/v2.0/token"
-        token = pes_url.split('/')[-1] == 'dummy'
+    data = {
+        "grant_type": "client_credentials",
+        "client_id": s.MS_ENTRA_CLIENT_ID,
+        "client_secret": s.MS_ENTRA_CLIENT_SECRET,
+        "scope": scope,
+    }
+
+    try:
+        response = requests.post(url, data=data)
+        response.raise_for_status()
+        return response.json()["access_token"]
+    except requests.exceptions.HTTPError as e:
+        logger.error(f"Could not fetch application token: {e}")
+        raise ValueError(f"Could not fetch application token!")
+
+
+def fetch_pes(request, pes_url, user) -> dict:
 
-    if token:
     for k, v in s.PES_API_MAPPING.items():
         if pes_url.startswith(k):
+
+            prod = "cropkey-np" not in pes_url
+
             pes_id = pes_url.split('/')[-1]
 
             if pes_id == 'dummy':
@@ -175,7 +209,16 @@ def fetch_pes(request, pes_url) -> dict:
                 res_data["pes_url"] = pes_url
                 return res_data
             else:
-                    headers = {"Authorization": f"Bearer {token['access_token']}"}
+                headers = {
+                    "accept": "*/*",
+                    "authorization": "Bearer " + get_application_token(prod),
+                }
+
+                # Restrict request if user is not part of any secret group
+                if not has_secret_group(user):
+                    headers["app-classification-level-restriction"] = "restrict-pes-secret-structure-access"
+
+
                 params = {"pes_reg_entity_corporate_id": pes_id}
 
                 res = requests.get(v, headers=headers, params=params, proxies=s.PROXIES or None)
@@ -184,6 +227,11 @@ def fetch_pes(request, pes_url) -> dict:
                     res.raise_for_status()
                     pes_data = res.json()
 
+                    # Handle missing response
+                    if "detail" in pes_data and "The following PES Reg Entities Corporate Ids could not be found" in pes_data["detail"]:
+                        raise ValueError(f"PES with id {pes_id} not found")
+
+                    # Ensure we have a entity
                     if len(pes_data) == 0:
                         raise ValueError(f"PES with id {pes_id} not found")
 
@@ -193,17 +241,15 @@ def fetch_pes(request, pes_url) -> dict:
 
                 except requests.exceptions.HTTPError as e:
                     raise ValueError(f"Error fetching PES with id {pes_id}: {e}")
-        else:
+
     raise ValueError(f"Unknown URL {pes_url}")
-    else:
-        raise ValueError("Could not fetch access token from request.")
 
 
 def visualize_pes(request):
     pes_link = request.GET.get('pesLink')
 
     if pes_link:
-        pes_data = fetch_pes(request, pes_link)
+        pes_data = fetch_pes(request, pes_link, request.user)
 
         representations = pes_data.get('representations')
 

bridge/contracts.py

@@ -1,4 +1,5 @@
 import enum
+from typing import Any, Dict
 from abc import ABC, abstractmethod
 
 from envipy_additional_information import EnviPyModel
@@ -69,6 +70,12 @@ class Plugin(ABC):
 
 
 class Property(Plugin):
+    def parameters(self) -> Dict[str, Any]:
+        """
+        Returns the parameters of the PropertyPlugin.
+        """
+        return {}
+
     @classmethod
     @abstractmethod
     def requires_rule_packages(cls) -> bool:
@@ -300,6 +307,12 @@ class Classifier(Plugin):
         """
         pass
 
+    def parameters(self) -> Dict[str, Any]:
+        """
+        Returns the parameters of the ClassifierPlugin.
+        """
+        return {}
+
     @abstractmethod
     def build(self, eP: EnviPyDTO, *args, **kwargs) -> BuildResult | None:
         """

entrypoint.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+set -e
+
+if [ "${SKIP_DJANGO_SETUP:-false}" != "true" ]; then
+  python manage.py migrate --no-input
+  python manage.py collectstatic --no-input
+fi
+
+exec "$@"

epapi/v1/interfaces/iuclid/projections.py

@@ -56,7 +56,7 @@ def get_pathway_for_iuclid_export(user, pathway_uuid: UUID) -> PathwayExportDTO:
 
         ai_for_node = []
         scenario_entries: list[PathwayScenarioDTO] = []
-        for scenario in sorted(node.scenarios.all(), key=lambda item: item.pk):
+        for scenario in sorted(node.get_scenarios(), key=lambda item: item.pk):
             ai_for_scenario = list(scenario.get_additional_information(direct_only=True))
             ai_for_node.extend(ai_for_scenario)
             scenario_entries.append(

epauth/views.py

@@ -9,7 +9,7 @@ from django.shortcuts import redirect
 
 from epdb.logic import UserManager, GroupManager
 from epdb.models import Group
-from epdb.views import get_remote_address
+from epdb.views import get_remote_address, error
 
 auth_log = logging.getLogger("auth")
 
@@ -72,6 +72,15 @@ def entra_callback(request):
 
     claims = result["id_token_claims"]
 
+    if claims.get("roles") is None or claims.get("roles") == [] or "envipath_registered_user" not in claims.get("roles"):
+        auth_log.error(f"Login attempt by {get_remote_address(request)} failed due to missing role")
+        return error(
+            request,
+            "Login Failed",
+            "The user is not authenticated. A reason for this might be a missing assignment to the respective enviPath group.",
+            403,
+        )
+
     user_name = claims.get("name")
     # preferred_username is a fallback for 2nd CWID
     user_email = claims.get("emailaddress", claims.get("email", claims.get("preferred_username")))
@@ -141,9 +150,9 @@ def entra_callback(request):
             if group.user_member.contains(u):
                 # User is eligible for secrete
                 pack = u.default_package
-                u.data_pool = group
+                pack.data_pool = group
-                u.classification_level = u.Classification.SECRET
+                pack.classification_level = pack.Classification.SECRET
-                u.save()
+                pack.save()
 
     # EDIT END
 

epdb/legacy_api.py

@@ -72,6 +72,10 @@ def get_cached_jwks(tenant_id: str, force=False) -> Dict:
     return jwks
 
 
+def get_package_for_read(user, package_uuid):
+    return PackageManager.get_package_by_id(user, package_uuid)
+
+
 def get_package_for_write(user, package_uuid):
     p = PackageManager.get_package_by_id(user, package_uuid)
     if not PackageManager.writable(user, p):
@@ -885,7 +889,7 @@ def create_package_compound(
             from bayer.models import PESCompound
 
             try:
-                pes_data = fetch_pes(request, c.pesLink)
+                pes_data = fetch_pes(request, c.pesLink, request.user)
             except ValueError as e:
                 return 400, {"message": f"Could not fetch PES data for {c.pesLink}"}
 
@@ -2010,7 +2014,7 @@ def add_pathway_node(request, package_uuid, pathway_uuid, n: Form[CreateNode]):
             from bayer.models import PESCompound
 
             try:
-                pes_data = fetch_pes(request, n.pesLink)
+                pes_data = fetch_pes(request, n.pesLink, request.user)
             except ValueError as e:
                 return 400, {"message": f"Could not fetch PES data for {n.pesLink}"}
 
@@ -2430,3 +2434,42 @@ def predict(request, np: Form[NonPersistent]):
         return 403, {
             "message": f"Getting Setting with id {np.setting_url} failed due to insufficient rights!"
         }
+
+
+##########
+# Export #
+##########
+class PackageExportInSchema(Schema):
+    package_uuid: str
+    additional_information_types: List[str] | None = None
+
+
+@router.get("/export", response={200: Any, 403: Error})
+def export(request, q: Query[PackageExportInSchema]):
+    try:
+        p = get_package_for_read(request.user, q.package_uuid)
+
+        from envipy_additional_information import registry
+        from utilities.misc import PathwayExporter
+
+        ai_types = []
+        if q.additional_information_types is not None:
+            for ai_type in q.additional_information_types:
+                if registry.get_model(ai_type) is None:
+                    return 400, {
+                        "message": f"Exporting Package with id {q.package_uuid} failed as {ai_type} is not a valid additional information type!"
+                    }
+                ai_types.append(ai_type)
+
+        exporter = PathwayExporter(p, add_infs_to_export=ai_types)
+        res = exporter.do_export()
+
+        filename = f"{p.get_name().replace(' ', '_')}_{p.uuid}.tsv"
+        response = HttpResponse(res, content_type="text/csv")
+        response["Content-Disposition"] = f'attachment; filename="{filename}"'
+
+        return response
+    except ValueError:
+        return 403, {
+            "message": f"Exporting Package with id {q.package_uuid} failed due to insufficient rights!"
+        }

epdb/logic.py

@@ -321,7 +321,7 @@ class GroupManager(object):
             else:
                 group.group_member.remove(member)
                 auth_log.info(
-                    f"{caller.username} ({caller.url}) removes {member.name} ({member.url}) to  {group.name} ({group.url})")
+                    f"{caller.username} ({caller.url}) removes {member.name} ({member.url}) from {group.name} ({group.url})")
         else:
             if add_or_remove == "add":
                 group.user_member.add(member)
@@ -329,7 +329,7 @@ class GroupManager(object):
             else:
                 group.user_member.remove(member)
                 auth_log.info(
-                    f"{caller.username} ({caller.url}) adds {member.username} ({member.url}) to  {group.name} ({group.url})")
+                    f"{caller.username} ({caller.url}) removes {member.username} ({member.url}) from {group.name} ({group.url})")
 
         group.save()
 

epdb/management/commands/localize_urls.py

@@ -44,15 +44,19 @@ class Command(BaseCommand):
             "EPModel",
             "ApplicabilityDomain",
             "EnzymeLink",
+            "AdditionalInformation",
         ]
         for model in MODELS:
             obj_cls = apps.get_model("epdb", model)
-            obj_cls.objects.update(
+
-                url=Replace(F("url"), Value(options["old"]), Value(options["new"]))
+            update_fields = {"url": Replace(F("url"), Value(options["old"]), Value(options["new"]))}
+            if hasattr(obj_cls, "description"):
+                update_fields["description"] = Replace(
+                    F("description"), Value(options["old"]), Value(options["new"])
                 )
+
             if issubclass(obj_cls, EnviPathModel):
-                obj_cls.objects.update(
+                update_fields["kv"] = Cast(
-                    kv=Cast(
                     Replace(
                         Cast(F("kv"), output_field=TextField()),
                         Value(options["old"]),
@@ -60,4 +64,5 @@ class Command(BaseCommand):
                     ),
                     output_field=JSONField(),
                 )
-                )
+
+            obj_cls.objects.update(**update_fields)

epdb/management/commands/reaction_rule_mapping.py

@@ -0,0 +1,97 @@
+import logging
+
+from django.core.management.base import BaseCommand
+from django.db import transaction
+from uuid import uuid4
+from epdb.models import Package, ReactionExplanation
+from utilities.chem import FormatConverter
+from django.utils import timezone
+
+
+logger = logging.getLogger(__name__)
+
+
+class Command(BaseCommand):
+    def add_arguments(self, parser):
+        parser.add_argument(
+            "--rule-package",
+            action="append",
+            default=["32de3cf4-e3e6-4168-956e-32fa5ddb0ce1"],
+            type=str,
+            help="UUID to process. Can be specified multiple times.",
+        )
+
+        parser.add_argument(
+            "--reaction-package",
+            action="append",
+            default=[
+                "32de3cf4-e3e6-4168-956e-32fa5ddb0ce1",  # BBD
+                "f05e38d8-e9b4-4c3e-b0d8-9ab29966eccf",  # Sediment
+                "521c547a-fd2a-491c-ad5b-7eaa1577fb65",  # Sludge
+                "5882df9c-dae1-4d80-a40e-db4724271456",  # Soil
+                "87a49584-d937-482c-9c33-25928dcb02a8",  # PFAS
+            ],
+            type=str,
+            help="UUID to process. Can be specified multiple times.",
+        )
+
+        parser.add_argument(
+            "--dry-run",
+            default=False,
+            action="store_true",
+            help="Perform dry run",
+        )
+
+    @transaction.atomic
+    def handle(self, *args, **options):
+        RUN_UUID = uuid4()
+        RUN_START = timezone.now()
+
+        rule_packages = Package.objects.filter(uuid__in=options["rule_package"])
+        reaction_packages = Package.objects.filter(uuid__in=options["reaction_package"])
+
+        rules = []
+        for rule_package in rule_packages:
+            rules.extend(rule_package.get_applicable_rules())
+
+        reactions = []
+        for reaction_package in reaction_packages:
+            reactions.extend(reaction_package.reactions)
+
+        logger.debug(f"Collected {len(rules)} rules and {len(reactions)} reactions.")
+
+        for i, reaction in enumerate(reactions):
+            logger.debug(f"Reaction {i} / {len(reactions)}")
+            for j, rule in enumerate(rules):
+                reactants, products = reaction.smirks().split(">>")
+
+                if len(reactants.split(".")) > 1:
+                    logger.debug(f"Skipping reaction {reaction.uuid} as it has multiple reactants.")
+                    break
+
+                products = products.split(".")
+
+                # Run reaction with rule
+                rule_products = rule.apply(reactants)
+
+                # Check if products match (in both directions if extras are not allowed)
+                for product_set in rule_products:
+                    covered, exact = FormatConverter.smiles_covered_by(
+                        products,
+                        product_set.product_set,
+                        standardize=True,
+                        canonicalize_tautomers=True,
+                        return_exact_match=True,
+                    )
+
+                    if covered and not options["dry-run"]:
+                        logger.debug(f"Reaction {reaction.uuid} explained by rule {rule.uuid}")
+                        re = ReactionExplanation()
+                        re.run_uuid = RUN_UUID
+                        re.run_start = RUN_START
+                        re.reaction = reaction
+                        re.rule = rule
+                        re.exact = exact
+                        re.save()
+                        # Its explained, if there are more sets skip them
+                        break

epdb/migrations/0028_auto_20260812_0902.py

@@ -0,0 +1,113 @@
+# Generated by Django 6.0.3 on 2026-08-12 09:02
+
+from django.conf import settings as s
+from django.db import migrations
+from envipy_additional_information import Likelihood, RuleLikelihood
+
+NEW_RULE = {
+    "parent": "bt0005",
+    "name": "bt0005-3667",
+    "description": "vic-unsubstituted Aromatic > vic-Dihydroxyaromatic",
+    "smirks": "[#8:7]([H])-[#6:1]([H])-1-[#6:2]=[#6:3]-[#6:4]=[#6:5]-[#6:6]([H])-1-[#8:8]([H])>>[#8:7]([H])-[#6:1]=1-[#6:2]=[#6:3]-[#6:4]=[#6:5]-[#6:6]=1-[#8:8]([H])",
+    "scenario_name": "bt0005-3667 aerobic likelihood",
+    "scenario_aerobic_likelihood": RuleLikelihood(likelihood=Likelihood.LIKELY),
+}
+
+RULE_FIXES = {
+    "bt0005-4282": "[c:1]([H])1:[c:2]([H]):[#6,#7;a:3]:[c:4]:[c:5]:[c:6]1>>[c:1]([#8])1:[c:2]([#8]):[#6,#7;a:3]:[c:4]:[c:5]:[c:6]1",
+    "bt0014-4215": "[c:1]([H])1[c:8][#6,#7;a:7][c:6][c:5][c:4]1[#8;!$([OH]c:[#6,#7;a:7]([OH])):9]([H])>>[#8:9]([H])[c:4]1:[c:5]:[c:6]:[#6,#7;a:7]:[c:8]:[c:1]1[#8]([H])",
+    # "bt0063-3938": "[#1,#6:6][#7;X3;!$(NC1CC1)!$([N][C]=O)!$([!#8]CNC=O):1]([#1,#6:7])[#6;A;X4:2][H:3]>>[#1,#6:6][#7;X3:1]([H:3])(=[#1,#6:7]).[#6;A:2]=O",
+    # CN1C=NC2=C1C(=O)N(C)C(=O)N2 not working anymore with bt0063-3938 if change above is applied
+    "bt0063-3938": "[#1,#6:6][#7;X3;!$(NC1CC1)!$([N][C]=O)!$([!#8]CNC=O):1]([#1,#6:7])[#6;A;X4:2][H:3]>>[#1,#6:6][#7;X3:1]([#1,#6:7])[H:3].[#6;A:2]=O",
+    "bt0068-3564": "[#7:4]!@-[#6:2](!@-[#7:1])=[O:5]>>[#7:4]-[#6:2](-[O+0H1])=[O:5].[#7H1:1]",
+    "bt0180-2844": "[H][C:2]([#6:5]([H])([H])([H]))([#1,#6:4])!@-[#6:1]([H])([H])-[#6:3](-[#8-:8])=[O:6]>>[#6:5]([H])([H])([H])\\[#6:2](-[#1,#6:4])=[#6H:1]\\[#6:3](-[#8-:8])=[O:6]",
+    "bt0181-1278": "[#8-:1]-[#6:2](=[O:11])-[#6:7]=[#6:8]-[#6:3](-[H])=[#6:5](-Cl)-[#6:6](-[#8-:10])=[O:9]>>[O+0H1:10]-[#6:6](=[O:9])-[#6:5]=[#6:3]-1-[O+0:1]-[#6:2](=[O:11])-[#6:7]=[#6:8]-1",
+    "bt0298-3335": "[#6:1][N+:2]#[C:3]>>[#6:1]-[#7H2:2]-[#6:3]=O",
+    "bt0322-3393": "[H:10]\\[#6:6](=[#6:9](/[#6:1]([H])([H])([H]))-[#6:11]-[#6:12]-[#6:13]=[#6:14])-[#6:5](-[#16:7])=[O:8]>>[H:10]\\[#6:6](-[#6:5](-[#16:7])=[O:8])=[#6:9](\\[#6:11]-[#6:12]-[#6:13]=[#6:14])-[#6:1]-[#6](-[#8-])=O",
+    "bt0343-2675": "[#8-]-[#6](=O)-[c:1]1[c:6][cH:7][c:8](-[#7H2,#8H1:9])[cH:10][c:11]1>>[#8H][c:1]1[c:6][c:7][c:8]([*:9])[c:10][c:11]1",
+    "bt0350-3319": "[#6:6][#7:3][#6;!R:2]=[#7;!R:1][#6:5]>>[#6:5][#7:1][#6:2]=O.[#6:6][#7:3]",  # Trig before 5 -> all of them shouldn't
+    "bt0374-4081": "[cH:4]1[c:16][c:15][c:14][c:13][c:3]1[#7,#8:2][c:1]1[c:8][c:9][c:10][c:11][c:12]1>>[#7,#8:2]-[c:1]1[c:12][c:11][c:10][c:9][c:8]1[c:13]1[c:14][c:15][c:16][c:4](-[#8])[c:3]1-[#8]",
+    "bt0378-3188": "[#8-:7][c:1]1[c:6]([#7+]([#8-])=O)[c:5][c:4]([#7+:9]([#8-])=O)[c:3][c:2]1([#7+:8]([#8-])=O)>>[#8+0:7]=[#6:1]1-[#6:6]-[#6:5]-[#6:4]([#7+:9]([#8-])=O)-[#6:3]-[#6:2]1([#7+:8]([#8-])=O)",
+    "bt0379-3190": "[#9,#17,#35,#53]-[#6:1](-[H])-1-[#6:5]-,=[#6:6]-[#6:7]-,=[#6:8]-[#6:2](-[H])-1-[#9,#17,#35,#53]>>[#6:6]~1-[#6:7]~[#6:8]-[#6:2]=[#6:1]-[#6:5]~1",
+    "bt0393-3367": "[#6:5]-[#6:1](-[#7:2](-[H])(-[H]))=[S+:3]-[#8-:6]>>[#6:5]-[#6:1](=[#7H1:2])-[S+0:3](=[O])-[#8+0H1:6]",
+}
+
+
+def forward_func(apps, schema_editor):
+    ContentType = apps.get_model("contenttypes", "ContentType")
+
+    pkg_class = s.EPDB_PACKAGE_MODEL
+
+    if len(pkg_class.split(".")) != 2:
+        raise ValueError(
+            f"EPDB_PACKAGE_MODEL must be of the form 'app_label.model_name', got {pkg_class}"
+        )
+
+    app_label, model_name = pkg_class.split(".")
+    Package = apps.get_model(app_label, model_name)
+    SimpleAmbitRule = apps.get_model("epdb", "SimpleAmbitRule")
+    ParallelRule = apps.get_model("epdb", "ParallelRule")
+    Scenario = apps.get_model("epdb", "Scenario")
+    AdditionalInformation = apps.get_model("epdb", "AdditionalInformation")
+    simple_ambit_rule_ct = ContentType.objects.get_for_model(SimpleAmbitRule)
+
+    if Package.objects.filter(name="EAWAG-BBD").exists():
+        p = Package.objects.get(name="EAWAG-BBD")
+
+        if not SimpleAmbitRule.objects.filter(package=p, name=NEW_RULE["name"]).exists():
+            # Create Missing Rule
+            new_sr = SimpleAmbitRule()
+            new_sr.polymorphic_ctype = simple_ambit_rule_ct
+            new_sr.package = p
+            new_sr.name = NEW_RULE["name"]
+            new_sr.description = NEW_RULE["description"]
+            new_sr.smirks = NEW_RULE["smirks"]
+            new_sr.save()
+
+            new_sr.url = "{}/simple-ambit-rule/{}".format(new_sr.package.url, new_sr.uuid)
+            new_sr.save()
+
+            # Add likelihood
+            new_scen = Scenario()
+            new_scen.package = p
+            new_scen.name = NEW_RULE["scenario_name"]
+            new_scen.save()
+
+            new_scen.url = "{}/scenario/{}".format(new_scen.package.url, new_scen.uuid)
+            new_scen.save()
+
+            ai = NEW_RULE["scenario_aerobic_likelihood"]
+            new_add_inf = AdditionalInformation()
+            new_add_inf.package = p
+            new_add_inf.type = ai.__class__.__name__
+            new_add_inf.data = ai.model_dump(mode="json")
+            new_add_inf.scenario = new_scen
+            new_add_inf.save()
+
+            new_add_inf.url = "{}/additional-information/{}".format(
+                new_add_inf.scenario.url, new_add_inf.uuid
+            )
+            new_add_inf.save()
+
+            # Link Scenario
+            new_sr.scenarios.add(new_scen)
+
+            # Link to bt0005
+            pr = ParallelRule.objects.get(package=p, name="bt0005")
+            pr.simple_rules.add(new_sr)
+
+        # Update others
+        for rule_name, smirks in RULE_FIXES.items():
+            sr = SimpleAmbitRule.objects.get(package=p, name=rule_name)
+            sr.smirks = smirks
+            sr.save()
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("epdb", "0027_alter_compound_aliases_and_more"),
+    ]
+
+    operations = [
+        migrations.RunPython(forward_func, reverse_code=migrations.RunPython.noop),
+    ]

epdb/migrations/0029_reactionexplanation_reaction_explained_by.py

@@ -0,0 +1,63 @@
+# Generated by Django 6.0.3 on 2026-08-13 09:58
+
+import django.db.models.deletion
+import django.utils.timezone
+import model_utils.fields
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("epdb", "0028_auto_20260812_0902"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="ReactionExplanation",
+            fields=[
+                (
+                    "id",
+                    models.BigAutoField(
+                        auto_created=True, primary_key=True, serialize=False, verbose_name="ID"
+                    ),
+                ),
+                (
+                    "created",
+                    model_utils.fields.AutoCreatedField(
+                        default=django.utils.timezone.now, editable=False, verbose_name="created"
+                    ),
+                ),
+                (
+                    "modified",
+                    model_utils.fields.AutoLastModifiedField(
+                        default=django.utils.timezone.now, editable=False, verbose_name="modified"
+                    ),
+                ),
+                ("run_uuid", models.UUIDField()),
+                ("run_start", models.DateTimeField()),
+                ("exact", models.BooleanField(default=False)),
+                (
+                    "reaction",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE, to="epdb.reaction"
+                    ),
+                ),
+                (
+                    "rule",
+                    models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, to="epdb.rule"),
+                ),
+            ],
+            options={
+                "abstract": False,
+            },
+        ),
+        migrations.AddField(
+            model_name="reaction",
+            name="explained_by",
+            field=models.ManyToManyField(
+                related_name="explained_reactions",
+                through="epdb.ReactionExplanation",
+                to="epdb.rule",
+            ),
+        ),
+    ]

epdb/migrations/0030_auto_20260814_0741.py

@@ -0,0 +1,37 @@
+# Generated by Django 6.0.3 on 2026-08-14 07:41
+
+from django.db import migrations
+
+
+def forward_func(apps, schema_editor):
+    ContentType = apps.get_model("contenttypes", "ContentType")
+    AdditionalInformation = apps.get_model("epdb", "AdditionalInformation")
+
+    models = {}
+
+    for c in ContentType.objects.all():
+        try:
+            models[(c.app_label, c.model)] = apps.get_model(c.app_label, c.model)
+        except Exception:
+            pass
+
+    for ai in AdditionalInformation.objects.all():
+        if ai.url is None:
+            if ai.content_type is None:
+                ai.url = "{}/additional-information/{}".format(ai.scenario.url, ai.uuid)
+            else:
+                model = models[(ai.content_type.app_label, ai.content_type.model)]
+                obj = model.objects.get(pk=ai.object_id)
+                ai.url = "{}/additional-information/{}".format(obj.url, ai.uuid)
+
+            ai.save()
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("epdb", "0029_reactionexplanation_reaction_explained_by"),
+    ]
+
+    operations = [
+        migrations.RunPython(forward_func, reverse_code=migrations.RunPython.noop),
+    ]

epdb/models.py

@@ -859,9 +859,15 @@ class Compound(
     @property
     def related_reactions(self):
         return (
+            (
                 Reaction.objects.filter(package=self.package, educts__in=[self.default_structure])
-            | Reaction.objects.filter(package=self.package, products__in=[self.default_structure])
+                | Reaction.objects.filter(
-        ).order_by("name")
+                    package=self.package, products__in=[self.default_structure]
+                )
+            )
+            .distinct()
+            .order_by("name")
+        )
 
     @property
     def related_nodes(self):
@@ -1735,6 +1741,14 @@ class SequentialRuleOrdering(models.Model):
     order_index = models.IntegerField(null=False, blank=False)
 
 
+class ReactionExplanation(TimeStampedModel):
+    run_uuid = models.UUIDField(null=False, blank=False)
+    run_start = models.DateTimeField(null=False, blank=False)
+    reaction = models.ForeignKey("epdb.Reaction", on_delete=models.CASCADE)
+    rule = models.ForeignKey("epdb.Rule", on_delete=models.CASCADE)
+    exact = models.BooleanField(default=False)
+
+
 class Reaction(
     EnviPathModel, AliasMixin, ScenarioMixin, ReactionIdentifierMixin, AdditionalInformationMixin
 ):
@@ -1760,6 +1774,12 @@ class Reaction(
 
     external_identifiers = GenericRelation("ExternalIdentifier")
 
+    explained_by = models.ManyToManyField(
+        "epdb.Rule",
+        through="ReactionExplanation",
+        related_name="explained_reactions",
+    )
+
     def _url(self):
         return "{}/reaction/{}".format(self.package.url, self.uuid)
 
@@ -2173,7 +2193,7 @@ class Pathway(EnviPathModel, AliasMixin, ScenarioMixin, AdditionalInformationMix
 
             row += [cs.smiles, cs.get_name(), n.depth]
 
-            edges = self.edges.filter(end_nodes__in=[n])
+            edges = self.edges.filter(end_nodes=n)
             if len(edges):
                 for e in edges:
                     _row = row.copy()
@@ -2585,7 +2605,7 @@ class Node(EnviPathModel, AliasMixin, ScenarioMixin, AdditionalInformationMixin)
                 collected[str(ai.scenario.uuid)]["proposed"] = True
 
             if ai.type == "Confidence":
-                collected[str(ai.scenario.uuid)]["Confidence"] = ai.get().level
+                collected[str(ai.scenario.uuid)]["Confidence"] = ai.get().level.value
 
             if ai.type == "TransformationProductImportance":
                 collected[str(ai.scenario.uuid)]["Transformation product importance"] = (
@@ -2815,21 +2835,6 @@ class PackageBasedModel(EPModel):
     )
     multigen_eval = models.BooleanField(null=False, blank=False, default=False)
 
-    def statistics(self):
-        return {
-            'accuracy': [
-                self.eval_results["average_accuracy"],
-                self.eval_results.get("multigen_average_accuracy")],
-            'precision': [
-                self.eval_results["average_precision_per_threshold"][f"{self.threshold:.2f}"],
-                self.eval_results.get("multigen_average_precision_per_threshold", {}).get(f"{self.threshold:.2f}")
-            ],
-            'recall': [
-                self.eval_results["average_recall_per_threshold"][f"{self.threshold:.2f}"],
-                self.eval_results.get("multigen_average_recall_per_threshold", {}).get(f"{self.threshold:.2f}")
-            ],
-        }
-
     @property
     def pr_curve(self):
         if self.model_status != self.FINISHED:
@@ -2873,6 +2878,58 @@ class PackageBasedModel(EPModel):
 
         return res
 
+    def parameters(self):
+        params = {
+            "Model Evaluation Threshold": f"{self.threshold:.2f}",
+            "Multi Gen Evaluation": "Yes" if self.multigen_eval else "No",
+        }
+
+        if self.app_domain:
+            params["Applicability Domain Num Neighbors"] = f"{self.app_domain.num_neighbours:.2f}"
+            params["Applicability Domain Reliability Threshold"] = (
+                f"{self.app_domain.reliability_threshold:.2f}"
+            )
+            params["Applicability Domain Local Compatibility Threshold"] = (
+                f"{self.app_domain.local_compatibilty_threshold:.2f}"
+            )
+
+        return params
+
+    def statistics(self):
+        from sklearn.metrics import auc
+
+        recall = list(self.eval_results["average_recall_per_threshold"].values())
+        precision = list(self.eval_results["average_precision_per_threshold"].values())
+        mg_recall = list(
+            self.eval_results.get("multigen_average_recall_per_threshold", {}).values()
+        )
+        mg_precision = list(
+            self.eval_results.get("multigen_average_precision_per_threshold", {}).values()
+        )
+
+        return {
+            "accuracy": [
+                self.eval_results["average_accuracy"],
+                self.eval_results.get("multigen_average_accuracy"),
+            ],
+            "precision": [
+                self.eval_results["average_precision_per_threshold"][f"{self.threshold:.2f}"],
+                self.eval_results.get("multigen_average_precision_per_threshold", {}).get(
+                    f"{self.threshold:.2f}"
+                ),
+            ],
+            "recall": [
+                self.eval_results["average_recall_per_threshold"][f"{self.threshold:.2f}"],
+                self.eval_results.get("multigen_average_recall_per_threshold", {}).get(
+                    f"{self.threshold:.2f}"
+                ),
+            ],
+            "Area under PR Curve": [
+                auc(recall, precision),
+                auc(mg_recall, mg_precision) if self.multigen_eval else None,
+            ],
+        }
+
     @cached_property
     def applicable_rules(self) -> List["Rule"]:
         """
@@ -3053,8 +3110,6 @@ class PackageBasedModel(EPModel):
                 thresholds.append(np.float64(threshold))
                 thresholds.sort()
 
-            logger.info(f"Thresholds: {thresholds}")
-
             precision = {f"{t:.2f}": [] for t in thresholds}
             recall = {f"{t:.2f}": [] for t in thresholds}
 
@@ -3103,14 +3158,17 @@ class PackageBasedModel(EPModel):
             for t in thresholds:
                 for true, pred in zip(pathways, pred_pathways):
                     acc, pre, rec = multigen_eval(true, pred, t)
-                    if abs(t - threshold) < 0.01:
+
-                        mg_acc = acc
+                    if f"{t:.2f}" == f"{threshold:.2f}":
+                        mg_acc += acc
+
                     precision[f"{t:.2f}"].append(pre)
                     recall[f"{t:.2f}"].append(rec)
 
+            avg_mg_acc = mg_acc / len(root_compounds)
             precision = {k: sum(v) / len(v) if len(v) > 0 else 0 for k, v in precision.items()}
             recall = {k: sum(v) / len(v) if len(v) > 0 else 0 for k, v in recall.items()}
-            return mg_acc, precision, recall
+            return avg_mg_acc, precision, recall
 
         # If there are eval packages perform single generation evaluation on them instead of random splits
         if self.eval_packages.count() > 0:
@@ -4242,6 +4300,9 @@ class ClassifierPluginModel(PackageBasedModel):
         instance = impl(conf)
         return instance
 
+    def parameters(self):
+        return self.instance().parameters()
+
     def build_dataset(self):
         """
         Required by general model contract but actual implementation resides in plugin.
@@ -4462,6 +4523,9 @@ class PropertyPluginModel(PackageBasedModel):
         instance = impl()
         return instance
 
+    def parameters(self):
+        return self.instance().parameters()
+
     def build_dataset(self):
         """
         Required by general model contract but actual implementation resides in plugin.

epdb/tasks.py

@@ -477,8 +477,7 @@ def batch_predict(
             limit=None,
             setting_overrides={
                 "max_nodes": num_tps,
-                "max_depth": num_tps,
+                "model_threshold": 0.0,
-                "model_threshold": 0.001,
             },
         )
 

epdb/views.py

@@ -1145,13 +1145,17 @@ def package_model(request, package_uuid, model_uuid):
                 for pr in pred_res:
                     if len(pr) > 0:
                         products = []
+
                         for prod_set in pr.product_sets:
                             logger.debug(f"Checking {prod_set}")
                             products.append(tuple([x for x in prod_set]))
 
+                        products = list(set(products))
+
+                        for prod in products:
                             res["pred"].append(
                                 {
-                                "products": list(set(products)),
+                                    "products": list(prod),
                                     "probability": pr.probability,
                                     "btrule": {k: getattr(pr.rule, k) for k in ["url", "name"]}
                                     if pr.rule is not None

epiuclid/serializers/pathway_mapper.py

@@ -351,7 +351,8 @@ class PathwayMapper:
 
         props = SoilPropertiesData()
 
-        for ai in ai_list:
+        for ai_obj in ai_list:
+            ai = ai_obj.get()
             if isinstance(ai, SoilTexture1) and props.soil_type is None:
                 props.soil_type = ai.type.value
             elif isinstance(ai, SoilTexture2):

templates/components/navbar.html

@@ -108,8 +108,7 @@
         {% elif not meta.url_contains_package and meta.user.default_package.get_classification_level_display == "Restricted" %}
           <img src="{% static 'images/restricted_mid.png' %}" width="200">
         {% elif not meta.url_contains_package and meta.user.default_package.get_classification_level_display == "Secret" %}
-          <img src="{% static 'images/secret_mid.png' %}" width="200">
+          <img src="{% static 'images/secret_mid.png' %}" width="120">
-
         {% endif %}
         {% if not public_mode %}
           <a id="search-trigger" role="button" class="cursor-pointer">

templates/objects/model/_model_base.html

@@ -117,6 +117,39 @@
           </div>
         </div>
       {% endif %}
+
+      {% if model.parameters %}
+        <!-- Model Parameters Panel -->
+        <div class="collapse-arrow bg-base-200 collapse">
+          <input type="checkbox" checked />
+          <div class="collapse-title text-xl font-medium">Model Parameters</div>
+          <div class="collapse-content">
+            <div class="flex justify-center">
+              <div
+                id="model-stats"
+                class="overflow-x-auto rounded-box shadow-md bg-base-100"
+              >
+                <table class="table table-fixed w-full">
+                  <thead class="text-base">
+                    <tr>
+                      <th class="w-3/5">Parameter</th>
+                      <th>Value</th>
+                    </tr>
+                  </thead>
+                  <tbody>
+                    {% for param, value in model.parameters.items %}
+                      <tr>
+                        <td>{{ param }}</td>
+                        <td>{{ value }}</td>
+                      </tr>
+                    {% endfor %}
+                  </tbody>
+                </table>
+              </div>
+            </div>
+          </div>
+        </div>
+      {% endif %}
     {% endblock %}
 
     {% block usemodel %}

templates/objects/model/classification_model.html

@@ -109,12 +109,12 @@
         res += "<th scope='row'>" + cnt + "</th>";
         res +=
           "<th scope='row'>" +
-          data[transformation]["products"][0].join(", ") +
+          data[transformation]["products"].join(", ") +
           "</th>";
         res +=
           "<th scope='row'>" +
           "<img width='400' src='{% url 'depict' %}?smiles=" +
-          encodeURIComponent(data[transformation]["products"][0].join(".")) +
+          encodeURIComponent(data[transformation]["products"].join(".")) +
           "'></th>";
         res +=
           "<th scope='row'>" +

templates/objects/node.html

@@ -139,7 +139,7 @@
                           </tr>
                           <tr>
                             <td>Model</td>
-                            <td>{{ half_lifes.0.model }}</td>
+                            <td>{{ half_lifes.0.model.value }}</td>
                           </tr>
                         </tbody>
                       </table>

templates/objects/pathway.html

@@ -537,7 +537,7 @@
         <input type="checkbox" />
         <div class="collapse-title text-xl font-medium">Setting</div>
         <div class="collapse-content">
-          {% with setting_to_render=pathway.setting can_be_default=False %}
+          {% with setting_to_render=pathway.setting_with_overrides can_be_default=False %}
             {% include "objects/setting_template.html" %}
           {% endwith %}
         </div>

templates/objects/user.html

@@ -95,6 +95,7 @@
     </div>
 
     <!-- Other Prediction Settings -->
+    {% if meta.available_settings|length > 1 %}
       <div class="collapse-arrow bg-base-200 collapse">
         <input type="checkbox" />
         <div class="collapse-title text-xl font-medium">
@@ -110,6 +111,7 @@
           {% endfor %}
         </div>
       </div>
+    {% endif %}
   </div>
 
   <script>

tests/test_jobs.py

@@ -58,7 +58,7 @@ class MultiGenTest(TestCase):
 
         self.assertEqual(
             pw.setting_with_overrides.max_depth,
-            f"{num_tps} (this is an override for this particular pathway)",
+            5,
         )
         self.assertEqual(
             pw.setting_with_overrides.max_nodes,

tests/views/test_model_views.py

@@ -47,7 +47,7 @@ class ModelViewTest(TestCase):
 
         expected = [
             {
-                "products": [["CCN(CC)C(=O)C1=CC(C=O)=CC=C1"]],
+                "products": ["CCN(CC)C(=O)C1=CC(C=O)=CC=C1"],
                 "probability": 0.75,
                 "btrule": {
                     "url": "http://localhost:8000/package/1869d3f0-60bb-41fd-b6f8-afa75ffb09d3/simple-ambit-rule/2f2e0c39-e109-4836-959f-2bda2524f022",
@@ -55,7 +55,7 @@ class ModelViewTest(TestCase):
                 },
             },
             {
-                "products": [["O=C(O)C1=CC(CO)=CC=C1", "CCNCC"]],
+                "products": ["O=C(O)C1=CC(CO)=CC=C1", "CCNCC"],
                 "probability": 0.25,
                 "btrule": {
                     "url": "http://localhost:8000/package/1869d3f0-60bb-41fd-b6f8-afa75ffb09d3/simple-ambit-rule/0e6e9290-b658-4450-b291-3ec19fa19206",
@@ -63,7 +63,7 @@ class ModelViewTest(TestCase):
                 },
             },
             {
-                "products": [["CCNC(=O)C1=CC(CO)=CC=C1", "CC=O"]],
+                "products": ["CCNC(=O)C1=CC(CO)=CC=C1", "CC=O"],
                 "probability": 0.0,
                 "btrule": {
                     "url": "http://localhost:8000/package/1869d3f0-60bb-41fd-b6f8-afa75ffb09d3/simple-ambit-rule/27a3a353-0b66-4228-bd16-e407949e90df",

utilities/chem.py

@@ -68,6 +68,8 @@ class PredictionResult(object):
 
 
 class FormatConverter(object):
+    tautomer_enumerator = rdMolStandardize.TautomerEnumerator()
+
     @staticmethod
     def mass(smiles):
         return Descriptors.MolWt(FormatConverter.from_smiles(smiles))
@@ -240,8 +242,9 @@ class FormatConverter(object):
             Chem.RemoveStereochemistry(res_mol)
 
         if canonicalize_tautomers:
-            te = rdMolStandardize.TautomerEnumerator()  # idem
+            tautomers = FormatConverter.tautomer_enumerator.Enumerate(res_mol)
-            res_mol = te.Canonicalize(res_mol)
+            if len(tautomers) >= 1:
+                res_mol = FormatConverter.tautomer_enumerator.PickCanonical(tautomers)
 
         return Chem.MolToSmiles(res_mol, kekuleSmiles=True)
 
@@ -389,7 +392,7 @@ class FormatConverter(object):
                                 prods.append(p)
 
                         except ValueError as e:
-                            logger.error(f"Sanitizing and converting failed:\n{e}")
+                            logger.debug(f"Sanitizing and converting failed:\n{e}")
                             continue
 
                     if len(prods):
@@ -397,7 +400,8 @@ class FormatConverter(object):
                         pss.add(ps)
 
         except Exception as e:
-            logger.error(f"Applying {smirks} on {smiles} failed:\n{e}")
+            logger.debug(f"Applying {smirks} on {smiles} failed:\n{e}")
+            pass
 
         return list(pss)
 
@@ -444,6 +448,7 @@ class FormatConverter(object):
         r_smiles: List[str],
         standardize: bool = True,
         canonicalize_tautomers: bool = True,
+        return_exact_match: bool = False,
     ) -> bool:
         """
         Check if all SMILES in the left list are covered by (contained in) the right list.
@@ -482,9 +487,11 @@ class FormatConverter(object):
         if standardize:
             for smi in l_smiles:
                 try:
-                    smi = FormatConverter.standardize(
+                    smi = FormatConverter.canonicalize(
+                        FormatConverter.standardize(
                             smi, remove_stereo=True, canonicalize_tautomers=canonicalize_tautomers
                         )
+                    )
                 except Exception:
                     # :shrug:
                     # logger.debug(f'Standardizing SMILES failed for {smi}')
@@ -497,8 +504,12 @@ class FormatConverter(object):
         if standardize:
             for smi in r_smiles:
                 try:
-                    smi = FormatConverter.standardize(
+                    smi = FormatConverter.canonicalize(
-                        smi, remove_stereo=True, canonicalize_tautomers=canonicalize_tautomers
+                        FormatConverter.standardize(
+                            smi,
+                            remove_stereo=True,
+                            canonicalize_tautomers=canonicalize_tautomers,
+                        )
                     )
                 except Exception:
                     # :shrug:
@@ -507,8 +518,11 @@ class FormatConverter(object):
                 standardized_r_smiles.append(smi)
         else:
             standardized_r_smiles = r_smiles
-
+        if not return_exact_match:
             return len(set(standardized_l_smiles).difference(set(standardized_r_smiles))) == 0
+        return len(set(standardized_l_smiles).difference(set(standardized_r_smiles))) == 0, set(
+            standardized_l_smiles
+        ) == set(standardized_r_smiles)
 
 
 class Standardizer(ABC):

utilities/misc.py

@@ -4,11 +4,13 @@ import hmac
 import json
 import logging
 import uuid
+from abc import ABC, abstractmethod
 from collections import defaultdict
 from datetime import datetime
-from typing import Any, Dict, List, Optional, TYPE_CHECKING
+from typing import Any, Dict, List, Optional, TYPE_CHECKING, Type
 
 from django.conf import settings as s
+from envipy_additional_information import EnviPyModel, UIConfig
 from ninja import Schema
 from pydantic import HttpUrl, ValidationError
 
@@ -156,7 +158,7 @@ class EnzymeExportSchema(RefEnzymeExportSchema):
 
 
 class EnzymeRuleExportSchema(RefRuleExportSchema):
-    enzymes: List[EnzymeExportSchema] | None = None
+    enzymes: List[EnzymeExportSchema] = []
 
     @staticmethod
     def resolve_enzymes(obj):
@@ -319,15 +321,20 @@ class PackageExportSchema(Schema):
         return AdditionalInformation.objects.filter(package=obj)
 
 
-class PackageExporter:
+class Exporter(ABC):
     def __init__(self, package: Package):
         self._raw_package = package
 
     def do_export(self):
-        return PackageExporter._export_package_as_json(self._raw_package)
+        return self._export()
 
-    @staticmethod
+    @abstractmethod
-    def _export_package_as_json(package: Package) -> Dict[str, Any]:
+    def _export(self):
+        pass
+
+
+class PackageExporter(Exporter):
+    def _export(self) -> Dict[str, Any]:
         """
         Dumps a Package and all its related objects as JSON.
 
@@ -338,11 +345,126 @@ class PackageExporter:
             Dict containing the complete package data as JSON-serializable structure
         """
 
-        data = PackageExportSchema.from_orm(package)
+        data = PackageExportSchema.from_orm(self._raw_package)
 
         return data.model_dump(mode="json")
 
 
+class PathwayExporter(Exporter):
+    def __init__(self, package: Package, add_infs_to_export: List[str] = []):
+        super().__init__(package)
+        self._add_infs_to_export = add_infs_to_export
+
+    def _flatten_additional_information(self, ai: AdditionalInformation) -> dict[str, Any]:
+        model_cls: Type[EnviPyModel] = type(ai.get())
+
+        def _flatten(d: dict, parent_key: str = "") -> dict[str, Any]:
+            items: dict[str, Any] = {}
+            for key, value in d.items():
+                new_key = f"{parent_key}__{key.lower()}" if parent_key else key.lower()
+                if isinstance(value, dict):
+                    items.update(_flatten(value, new_key))
+                else:
+                    items[new_key] = value
+            return items
+
+        flat = _flatten(ai.data, ai.type)
+
+        ui_class = getattr(model_cls, "UI", None)
+        if ui_class is None:
+            return flat
+
+        for f in model_cls.model_fields:
+            ui_info = getattr(ui_class, f, None)
+
+            if not isinstance(ui_info, UIConfig) or ui_info.unit is None:
+                continue
+
+            flat[f"{model_cls.__name__}__{f}__unit"] = ui_info.unit
+
+        return flat
+
+    def _export(self):
+        from io import StringIO
+        from csv import DictWriter
+
+        rows = []
+
+        for pw in self._raw_package.pathways.all():
+            for n in pw.nodes:
+                for scen in pw.scenarios.all():
+                    row = {
+                        "pathway_name": pw.name,
+                        "pathway_id": str(pw.url),
+                        "node_depth": n.depth,
+                        "compound_id": str(n.default_node_label.compound.url),
+                        "pubchem_ID": n.default_node_label.pubchem_compound_id,
+                        "compound_name": n.default_node_label.compound.name,
+                        "compound_smiles": n.default_node_label.smiles,
+                        "scenario_id": str(scen.url),
+                        "scenario_name": scen.name,
+                        "scenario_type": scen.scenario_type,
+                        "scenario_description": scen.description,
+                    }
+
+                    if self._add_infs_to_export:
+                        ai_qs = AdditionalInformation.objects.filter(
+                            scenario=scen, type__in=self._add_infs_to_export
+                        )
+                    else:
+                        ai_qs = AdditionalInformation.objects.filter(scenario=scen)
+
+                    for ai in ai_qs:
+                        if ai.type == "ProposedIntermediate" and ai.content_object == n:
+                            row.update({"proposed_intermediate": True})
+                        elif ai.type == "SpikeCompound":
+                            spike = {"SpikeCompound__url": ai.get().url}
+
+                            try:
+                                struc = CompoundStructure.objects.get(
+                                    compound__package=self._raw_package, url=ai.get().url
+                                )
+                                spike["SpikeCompound__smiles"] = struc.smiles
+                            except Exception:
+                                spike["SpikeCompound__smiles"] = None
+
+                            row.update(**spike)
+                        else:
+                            row.update(self._flatten_additional_information(ai))
+                    rows.append(row)
+
+        # Get all header fields
+        all_header_fields = set()
+        for row in rows:
+            all_header_fields.update(row.keys())
+
+        # Per request the CSV should start with these fields
+        header = [
+            "pathway_name",
+            "pathway_id",
+            "node_depth",
+            "compound_id",
+            "pubchem_ID",
+            "compound_name",
+            "compound_smiles",
+            "scenario_id",
+            "scenario_name",
+            "scenario_type",
+            "scenario_description",
+        ]
+
+        # User remaining fields and place them after the predefined values in a sorted manner
+        remainder = sorted(list(all_header_fields.difference(set(header))))
+        header.extend(remainder)
+
+        buffer = StringIO()
+        writer = DictWriter(buffer, fieldnames=header, delimiter="\t")
+        writer.writeheader()
+        writer.writerows(rows)
+        buffer.seek(0)
+        return buffer.getvalue()
+
+
 class PackageImporter:
     def __init__(self, package: Dict[str, Any], preserve_uuids: bool = False):
         self.preserve_uuids = preserve_uuids

utilities/ml.py

@@ -238,9 +238,11 @@ class RuleBasedDataset(Dataset):
     ):
         if feat_funcs is None:
             feat_funcs = [FormatConverter.maccs]
-        _structures = set()  # Get all the structures
+
+        _structures = set()
         for r in reactions:
             _structures.update(r.educts.all())
+
             if not educts_only:
                 _structures.update(r.products.all())
 
@@ -282,17 +284,14 @@ class RuleBasedDataset(Dataset):
                     if key not in triggered:
                         continue
 
-                    # standardize products from reactions for comparison
+                    if FormatConverter.smiles_covered_by(
-                    standardized_products = []
+                        [prod.smiles for prod in r.products.all()],
-                    for cs in r.products.all():
+                        list(triggered[key]),
-                        smi = cs.smiles
+                        standardize=True,
-                        try:
+                        canonicalize_tautomers=True,
-                            smi = FormatConverter.standardize(smi, remove_stereo=True)
+                    ):
-                        except Exception:
-                            logger.debug(f"Standardizing SMILES failed for {smi}")
-                        standardized_products.append(smi)
-                    if len(set(standardized_products).difference(triggered[key])) == 0:
                         observed.add(key)
+
         feat_columns = []
         for feat_func in feat_funcs:
             if isinstance(feat_func, Descriptor):
@@ -301,6 +300,7 @@ class RuleBasedDataset(Dataset):
                 feats = feat_func(compounds[0].smiles)
             start_i = len(feat_columns)
             feat_columns.extend([f"feature_{start_i + i}" for i, _ in enumerate(feats)])
+
         ds_columns = (
             ["structure_id"]
             + feat_columns
@@ -334,7 +334,9 @@ class RuleBasedDataset(Dataset):
                     obs.append(None)
                 else:
                     obs.append(0)
+
             rows.append([str(comp.uuid)] + feats + trig + obs)
+
         ds = RuleBasedDataset(len(applicable_rules), ds_columns, data=rows)
         return ds
 
@@ -680,9 +682,13 @@ class RelativeReasoning:
 
     def predict(self, X):
         res = np.zeros((len(X), (self.end_index + 1 - self.start_index)))
+        immutable_res = np.zeros((len(X), (self.end_index + 1 - self.start_index)))
 
         # Loop through all instances
         for inst_idx, inst in enumerate(X):
+            for i, t in enumerate(inst[self.start_index : self.end_index + 1]):
+                immutable_res[inst_idx][i] = t
+
             # Loop through all "triggered" features
             for i, t in enumerate(inst[self.start_index : self.end_index + 1]):
                 # Set label
@@ -696,7 +702,7 @@ class RelativeReasoning:
                             if i2 in self.winmap.get(i, []):
                                 # if thatat rule also triggered, it dominated the current
                                 # set label to 0
-                                if X[inst_idx][i2]:
+                                if immutable_res[inst_idx][i2]:
                                     res[inst_idx][i] = 0
 
         return res