Merge pull request #361 from bmeg/rc5-patch

Fix G2P -> Publication links from DGIGB
bmeg · Dec 23, 2019 · 3ec2706 · 3ec2706
2 parents e57def7 + a80511b
commit 3ec2706
Show file tree

Hide file tree

Showing 10 changed files with 76 additions and 53 deletions.
diff --git a/outputs.bmeg_manifest.dvc b/outputs.bmeg_manifest.dvc
@@ -1,4 +1,4 @@
-md5: 8799729d258acd405bdf6d92de68dfcf
+md5: fa537571e66fb976ab6f547028ddf408
 cmd: echo generating file manifest...
 deps:
 - md5: b2d2ab73b01758f77505caecc3963f29
@@ -69,23 +69,23 @@ deps:
   path: outputs/ccle/maf.SomaticCallset_Aliquots_Aliquot.Edge.json.gz
 - md5: 4dfc12d1a89eee58e38ef9109bbd64e4
   path: outputs/celllines/Case_SameAs_Case.Edge.json.gz
-- md5: 8e8e11059b5cbd3c1ae85573b30dabb0
+- md5: 015143f9421a9fb5a3311f92043056c2
   path: outputs/compound/normalized.Case_Compounds_Compound.Edge.json.gz
-- md5: dd0ac5a944094d76a0cf12c4bff575da
+- md5: 20b8a955bff703c00690081bebfbd372
   path: outputs/compound/normalized.Compound.Vertex.json.gz
-- md5: 488eee20c9e4f0c99070eb6dfbb3fdb7
+- md5: acb155e76510e2661158e52272338a3c
   path: outputs/compound/normalized.Compound_Cases_Case.Edge.json.gz
-- md5: 967a3e45c1ea2270a9b9bb4763b9cff9
+- md5: a0af3202a4dcef822bd16127407c043e
   path: outputs/compound/normalized.Compound_DrugResponses_DrugResponse.Edge.json.gz
-- md5: 21734e88b95e98c4eabac40c91cfbd45
+- md5: 6a029d0ee0b8f531cce4410d0e8affa4
   path: outputs/compound/normalized.Compound_G2PAssociations_G2PAssociation.Edge.json.gz
-- md5: a9945dd00371f9ae553098b2b71f4ce8
+- md5: d8cfbd16c89e661cf15df5c37cea5d72
   path: outputs/compound/normalized.Compound_Projects_Project.Edge.json.gz
-- md5: fb5f300161494a94d2493cfb5b704e68
+- md5: 645c674a623e27bc84f9449ad8111683
   path: outputs/compound/normalized.DrugResponse_Compounds_Compound.Edge.json.gz
-- md5: c76490317a7a49562e903c52ddba2c6e
+- md5: 69b9e6f4ee8226dac032bf8814d72ff2
   path: outputs/compound/normalized.G2PAssociation_Compounds_Compound.Edge.json.gz
-- md5: c53f52b19e36799ee099edc0ef3d7657
+- md5: a56821071854cc805120007a99728f0e
   path: outputs/compound/normalized.Project_Compounds_Compound.Edge.json.gz
 - md5: cb3fcbfcee28d6434f8397355cff568a
   path: outputs/ctrp/ctrp.Aliquot.Vertex.json.gz
@@ -113,15 +113,15 @@ deps:
   path: outputs/ctrp/ctrp.Sample_Aliquots_Aliquot.Edge.json.gz
 - md5: ce514bf610e31c6b36b3d2945d2e9ccc
   path: outputs/ctrp/ctrp.Sample_Case_Case.Edge.json.gz
-- md5: 3fa57b7c369e2c8dd2dc0d0b6c44766e
+- md5: aac51ae53ff021ed9291b1aa2ae78076
   path: outputs/dgidb/G2PAssociation.Vertex.json.gz
-- md5: 166177ce31138f0ee32dc339fdfdd987
+- md5: 758362f57be6a69c6969a48182e0ff18
   path: outputs/dgidb/G2PAssociation_Genes_Gene.Edge.json.gz
-- md5: 4f99308481017576d6cc208b6f514eab
+- md5: 57d481f7135e3a8761ad8920ceca5fb0
   path: outputs/dgidb/G2PAssociation_Publications_Publication.Edge.json.gz
-- md5: 97282e6b782ab81889884ba6d9ba368e
+- md5: a701cee66767a702c18b7b3d4cb4b829
   path: outputs/dgidb/Gene_G2PAssociations_G2PAssociation.Edge.json.gz
-- md5: 03f13f3513d5282b8a1f309429ae81f0
+- md5: cecbf1a9f863ecfd0ce3db553becfad9
   path: outputs/dgidb/Publication_G2PAssociations_G2PAssociation.Edge.json.gz
 - md5: 3f89bd640983b7ea76b5dbbbcd846941
   path: outputs/ensembl/Exon.Vertex.json.gz
@@ -359,7 +359,7 @@ deps:
   path: outputs/phenotype/normalized.Phenotype_Samples_Sample.Edge.json.gz
 - md5: 3729d0155dbf3f3f30b926a030a1cf95
   path: outputs/phenotype/normalized.Sample_Phenotypes_Phenotype.Edge.json.gz
-- md5: 87182b1ccb82bc03a2efb61ca77703d5
+- md5: 7bbe3da8a9ade7327c41858d20107e58
   path: outputs/publication/stub.Publication.Vertex.json.gz
 - md5: 6bdd6ef5f03bc16023a4c59bfcec95db
   path: outputs/pubmed/baseline/pubmed19n0001.Publication.Vertex.json.gz

diff --git a/outputs/compound/normalized.compounds.dvc b/outputs/compound/normalized.compounds.dvc
@@ -1,4 +1,4 @@
-md5: 72113976b51fe192773b38a2d2071503
+md5: a8590f84b821452a9ac88e92ab816e88
 cmd: python3 transform/compound/transform.py
 wdir: ../..
 deps:
@@ -14,7 +14,7 @@ deps:
   path: outputs/gdc/gdc.Compound.Vertex.json.gz
 - md5: bdd1c3dedf67c2f3179f04771115a3c7
   path: outputs/pharmacodb/Compound.Vertex.json.gz
-- md5: f7a2bbfb7436d76e09b9e13bd3d0c9ff
+- md5: 502320cf8546d8607d62a1f0518e97cc
   path: outputs/dgidb/Compound.Vertex.json.gz
 - md5: 90070ff6332de4a429188a69ea8d9633
   path: outputs/pharmacodb/DrugResponse_Compounds_Compound.Edge.json.gz
@@ -36,52 +36,52 @@ deps:
   path: outputs/pharmacodb/Project_Compounds_Compound.Edge.json.gz
 - md5: 62990f11a0051b7357506699167a7611
   path: outputs/pharmacodb/Compound_Projects_Project.Edge.json.gz
-- md5: 9c4627785660215fa9adbf7c9f6cf172
+- md5: 708a48991dc4d6051097e6a8e8175b7b
   path: outputs/dgidb/G2PAssociation_Compounds_Compound.Edge.json.gz
-- md5: fa0ae005deb28af2d935cc818c1394b4
+- md5: 0e9b6635d4e8153d221d1c2a0604627a
   path: outputs/dgidb/Compound_G2PAssociations_G2PAssociation.Edge.json.gz
 outs:
-- md5: dd0ac5a944094d76a0cf12c4bff575da
+- md5: 20b8a955bff703c00690081bebfbd372
   path: outputs/compound/normalized.Compound.Vertex.json.gz
   cache: true
   metric: false
   persist: false
-- md5: fb5f300161494a94d2493cfb5b704e68
+- md5: 645c674a623e27bc84f9449ad8111683
   path: outputs/compound/normalized.DrugResponse_Compounds_Compound.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 967a3e45c1ea2270a9b9bb4763b9cff9
+- md5: a0af3202a4dcef822bd16127407c043e
   path: outputs/compound/normalized.Compound_DrugResponses_DrugResponse.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: c53f52b19e36799ee099edc0ef3d7657
+- md5: a56821071854cc805120007a99728f0e
   path: outputs/compound/normalized.Project_Compounds_Compound.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: a9945dd00371f9ae553098b2b71f4ce8
+- md5: d8cfbd16c89e661cf15df5c37cea5d72
   path: outputs/compound/normalized.Compound_Projects_Project.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 488eee20c9e4f0c99070eb6dfbb3fdb7
+- md5: acb155e76510e2661158e52272338a3c
   path: outputs/compound/normalized.Compound_Cases_Case.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 8e8e11059b5cbd3c1ae85573b30dabb0
+- md5: 015143f9421a9fb5a3311f92043056c2
   path: outputs/compound/normalized.Case_Compounds_Compound.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: c76490317a7a49562e903c52ddba2c6e
+- md5: 69b9e6f4ee8226dac032bf8814d72ff2
   path: outputs/compound/normalized.G2PAssociation_Compounds_Compound.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 21734e88b95e98c4eabac40c91cfbd45
+- md5: 6a029d0ee0b8f531cce4410d0e8affa4
   path: outputs/compound/normalized.Compound_G2PAssociations_G2PAssociation.Edge.json.gz
   cache: true
   metric: false

diff --git a/outputs/dgidb/dgidb.dvc b/outputs/dgidb/dgidb.dvc
@@ -1,4 +1,4 @@
-md5: e720598cc16fd7bb5918bd09f0397589
+md5: 34c1bbfdb0868d369c4283b8982f2137
 cmd: python3 transform/dgidb/transform.py
 wdir: ../..
 deps:
@@ -8,45 +8,45 @@ deps:
   path: source/drug_enricher/drug_alias.tsv
 - md5: 64e7a82c87e7151a7c49846469157547
   path: src/bmeg/enrichers/drug_enricher.py
-- md5: 936873de1ea7a748293263ff126a2967
+- md5: 1f1f691fa223e6c7c05cb9324a68d96a
   path: transform/dgidb/transform.py
 outs:
-- md5: 3fa57b7c369e2c8dd2dc0d0b6c44766e
+- md5: aac51ae53ff021ed9291b1aa2ae78076
   path: outputs/dgidb/G2PAssociation.Vertex.json.gz
   cache: true
   metric: false
   persist: false
-- md5: f7a2bbfb7436d76e09b9e13bd3d0c9ff
+- md5: 502320cf8546d8607d62a1f0518e97cc
   path: outputs/dgidb/Compound.Vertex.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 166177ce31138f0ee32dc339fdfdd987
+- md5: 758362f57be6a69c6969a48182e0ff18
   path: outputs/dgidb/G2PAssociation_Genes_Gene.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 4f99308481017576d6cc208b6f514eab
+- md5: 57d481f7135e3a8761ad8920ceca5fb0
   path: outputs/dgidb/G2PAssociation_Publications_Publication.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 9c4627785660215fa9adbf7c9f6cf172
+- md5: 708a48991dc4d6051097e6a8e8175b7b
   path: outputs/dgidb/G2PAssociation_Compounds_Compound.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 03f13f3513d5282b8a1f309429ae81f0
+- md5: cecbf1a9f863ecfd0ce3db553becfad9
   path: outputs/dgidb/Publication_G2PAssociations_G2PAssociation.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: 97282e6b782ab81889884ba6d9ba368e
+- md5: a701cee66767a702c18b7b3d4cb4b829
   path: outputs/dgidb/Gene_G2PAssociations_G2PAssociation.Edge.json.gz
   cache: true
   metric: false
   persist: false
-- md5: fa0ae005deb28af2d935cc818c1394b4
+- md5: 0e9b6635d4e8153d221d1c2a0604627a
   path: outputs/dgidb/Compound_G2PAssociations_G2PAssociation.Edge.json.gz
   cache: true
   metric: false

diff --git a/outputs/publication/stub_publications.dvc b/outputs/publication/stub_publications.dvc
@@ -1,16 +1,16 @@
-md5: 9c3bbc179474e91c7ae2da409b34ca88
+md5: b76b1becf88a1b639c81da86933d69dc
 cmd: python3 transform/publication/transform.py
 wdir: ../..
 deps:
-- md5: 15fc6631cd700953dc7bcc74728c44fa
+- md5: ba4e56706b22e7572ef4591a999a5a58
   path: transform/publication/transform.py
 - md5: 1bef5c26bd8aab5b43de9c6e9f8d87de
   path: outputs/g2p/G2PAssociation_Publications_Publication.Edge.json.gz
 - md5: f2b1f26f98bc1f94fa528cbafbb6c60a
   path: outputs/g2p/Publication_G2PAssociations_G2PAssociation.Edge.json.gz
-- md5: 4f99308481017576d6cc208b6f514eab
+- md5: 57d481f7135e3a8761ad8920ceca5fb0
   path: outputs/dgidb/G2PAssociation_Publications_Publication.Edge.json.gz
-- md5: 03f13f3513d5282b8a1f309429ae81f0
+- md5: cecbf1a9f863ecfd0ce3db553becfad9
   path: outputs/dgidb/Publication_G2PAssociations_G2PAssociation.Edge.json.gz
 - md5: 2a69a3d9fe9acd1c5465afc03be20cd2.dir
   path: outputs/pubmed/baseline
@@ -23,7 +23,7 @@ deps:
 - md5: 737bc587e7faef80a008f30b9f310643
   path: outputs/pathway_commons/Interaction_Publications_Publication.Edge.json.gz
 outs:
-- md5: 87182b1ccb82bc03a2efb61ca77703d5
+- md5: 7bbe3da8a9ade7327c41858d20107e58
   path: outputs/publication/stub.Publication.Vertex.json.gz
   cache: true
   metric: false

diff --git a/src/bmeg/bmeg-dictionary b/src/bmeg/bmeg-dictionary
diff --git a/tests/unit/dgidb/source/dgidb/interactions.tsv b/tests/unit/dgidb/source/dgidb/interactions.tsv
@@ -8,3 +8,4 @@ PDGFRB	PDGFR	5159	ChemblInteractions	inhibitor	CHEMBL576982	QUIZARTINIB	QUIZARTI
 MTOR	2109	2475	GuideToPharmacologyInteractions	inhibitor	9361	WYE-354	CHEMBL561708	CHEMBL561708	
 DNTT	DNTT	1791	NCI		HYDROXYUREA	HYDROXYUREA	HYDROXYUREA	CHEMBL467	291471
 OPRM1	319	4988	GuideToPharmacologyInteractions	agonist	3534	SUFENTANIL	SUFENTANIL	CHEMBL658	
+PDGFRA	PDGFRA	5156	DoCM		IMATINIB	IMATINIB	IMATINIB	CHEMBL941	15928335,15685537,22718859,16638875,15146165,12949711,26130666,25157968,24132921,14645423,16954519,18794084,22745105
diff --git a/tests/unit/dgidb/test_dgidb_transform.py b/tests/unit/dgidb/test_dgidb_transform.py
@@ -2,7 +2,10 @@
 import contextlib
 import pytest
 import shutil
+import json
+
 from transform.dgidb.transform import transform
+from bmeg.ioutils import reader
 
 
 @pytest.fixture
@@ -50,6 +53,14 @@ def validate(helpers, interactions_file, emitter_directory):
         exclude_labels=['Publication', 'Gene']
     )
 
+    count = 0
+    with reader(pub_edge_file) as f:
+        for line in f:
+            e = json.loads(line)
+            assert e['to'] != 'Publication:ncbi.nlm.nih.gov/pubmed/'
+            count += 1
+    assert count == 16
+
 
 def test_simple(helpers, interactions_file, emitter_directory):
     """ simple test """

diff --git a/transform/dgidb/transform.py b/transform/dgidb/transform.py
@@ -94,7 +94,7 @@ def transform(interactions_file="source/dgidb/interactions.tsv",
             ),
             emit_backref=True
         )
-        if line["PMIDs"] is None or line["PMIDs"] != "":
+        if line["PMIDs"] is None or line["PMIDs"] == "":
             continue
         pubs = line["PMIDs"].split(",")
         for p in pubs:

diff --git a/transform/ensembl/uniprot.py b/transform/ensembl/uniprot.py
@@ -38,6 +38,7 @@ def transform(protein_table_path='source/ensembl/Homo_sapiens.GRCh37.85.uniprot.
 
         if uniprot_id not in emitted:
             p = Uniprot(id=Uniprot.make_gid(uniprot_id),
+                        uniprot_id=uniprot_id,
                         genome=GENOME_BUILD,
                         project_id=PROJECT_ID)
             emitter.emit_vertex(p)

diff --git a/transform/publication/transform.py b/transform/publication/transform.py
@@ -50,33 +50,43 @@ def transform(
     e = f = r = 0
     for file in files:
         f += 1
-        logging.info("processing HasSupportingReference file: {}/{}".format(f, nfiles))
+        logging.info("processing file: {}/{}".format(f, nfiles))
         with reader(file) as ins:
             for line in ins:
                 try:
                     edge = ujson.loads(line)
-                    if 'Publication:' not in edge['gid']:
+                    pid = None
+                    # get edge components
+                    if edge['to'].startswith('Publication'):
+                        pid = edge['to']
+                    elif edge['from'].startswith('Publication'):
+                        pid = edge['from']
+                    else:
                         logging.info('Edge {} has no publications that need transformation. skipping.'.format(file))
                         break
-                    # get edge components
-                    to = edge['to']
-                    if to in dedup:
+
+                    if pid in dedup:
                         r += 1
                         continue
-                    dedup[to] = True
-                    url = to.replace('Publication:', 'http://')
+
+                    url = pid.replace('Publication:', 'http://')
                     publication = Publication(
                         id=Publication.make_gid(url),
                         url=url,
                         project_id=Project.make_gid("Reference")
                     )
                     emitter.emit_vertex(publication)
+
+                    dedup[pid] = True
                     e += 1
+
                 except Exception as exc:
                     logging.error(str(exc))
                     raise exc
+
                 if e % batch_size == 0:
                     logging.info('emitted stub publication vertices: {}'.format(e))
+
     logging.info('emitted stub publication vertices: {}'.format(e))
     logging.info('existing publication refs found: {}'.format(r))
     emitter.close()