From cfed08b9bc56ea8837b5f63194da142f148fdd4e Mon Sep 17 00:00:00 2001
From: AndiMajore <andi.majore@googlemail.com>
Date: Fri, 5 May 2023 16:49:08 +0200
Subject: [PATCH] fixing drug-target-interaction update routine
Former-commit-id: 0b2f1a7cbbc8478e6d464b65a75bfa2e799f9d05 [formerly 76c7bea1b462da0f1eebf7d16367320e3eeffa56]
Former-commit-id: 009e32b1d3120a54f4088f5e073d264aa86ac29f
---
.../management/commands/import_from_nedrex.py | 15 ++++++++++++---
scripts/docker-entrypoint.sh | 1 -
2 files changed, 12 insertions(+), 4 deletions(-)
diff --git a/drugstone/management/commands/import_from_nedrex.py b/drugstone/management/commands/import_from_nedrex.py
index a307523..642a330 100644
--- a/drugstone/management/commands/import_from_nedrex.py
+++ b/drugstone/management/commands/import_from_nedrex.py
@@ -103,7 +103,7 @@ class NedrexImporter:
adjusted_name = name[:idx1 - 1].strip() if idx1 > 0 else ''
if '=' in adjusted_name:
idx2 = adjusted_name.index('=')
- return adjusted_name[idx2+1:].strip()
+ return adjusted_name[idx2 + 1:].strip()
return adjusted_name
return name
@@ -224,10 +224,11 @@ class NedrexImporter:
self.cache.init_proteins()
bulk = set()
- existing = set()
+ delete = set()
+ existing = dict()
if update:
for edge in models.ProteinDrugInteraction.objects.filter(pdi_dataset=dataset):
- existing.add(edge.__hash__())
+ existing[edge.__hash__()] = edge
source_datasets = dict()
source_is_licenced = dict()
@@ -248,6 +249,12 @@ class NedrexImporter:
protein = self.cache.get_protein_by_uniprot(to_id(edge['targetDomainId']))
actions = json.dumps(edge['actions'])
e = models.ProteinDrugInteraction(pdi_dataset=dataset, drug=drug, protein=protein, actions=actions)
+ if update and e.__hash__() in existing:
+ if existing[e.__hash__()] != e:
+ delete.add(existing[e.__hash__()])
+ del existing[e.__hash__()]
+ else:
+ return
if not update or e.__hash__() not in existing:
bulk.add(e)
for source in edge['dataSources']:
@@ -260,6 +267,8 @@ class NedrexImporter:
pass
iter_edge_collection('drug_has_target', add_dpi)
+ for d in delete:
+ d.delete()
models.ProteinDrugInteraction.objects.bulk_create(bulk)
# new_datasets = [dataset].extend(source_datasets.values())
# DatasetLoader.remove_old_pdi_data(new_datasets, licenced)
diff --git a/scripts/docker-entrypoint.sh b/scripts/docker-entrypoint.sh
index 4b187a4..079cbad 100755
--- a/scripts/docker-entrypoint.sh
+++ b/scripts/docker-entrypoint.sh
@@ -8,7 +8,6 @@ if [ -z "$DB_UPDATE_ON_START" ] || [ "$DB_UPDATE_ON_START" = "0" ]
then
echo "Update on startup disabled!"
else
- python3 manage.py populate_db --delete_model PDI
python3 manage.py populate_db --update -a
python3 manage.py make_graphs
fi
--
GitLab