nlp-kg/clinical_trial_graph

COVID-19 Trial Graph

This repositry contains scripts and data files for manuscript "COVID-19 Trial Graph: A Linked Graph for COVID-19 Clinical Trials".

Neo4j database import

Please download and install Neo4j Desktop from https://neo4j.com/download/ . All the data files (in csv format) for COVID-19 Trial Graph are provided in the neo4j_data folder. To import the data to Neo4j database, please do the following steps:

1. Create a local database in Neo4j desktop and create your own user and password, which will be used to connect the database and load your data
2. Activate the database by clicking “Start” button.
3. Click three-dot button in your database and choose “manage”. Once the new page shows up, click “Open Folder” button, where you are going to put the csv data files
4. Copy neo4j_data folder into database’s import directory
5. Replace the user and password in csv_to_neo4j.py and run the script

Neo4j Python Driver is required:

pip install neo4j

Screenshot of COVID-19 Trial Graph

Cypher query evaluation

Case query 1

Retrieve all COVID-19 clinical trials that target “remdesivir” as the intervention

match (c:Clinicaltrial)-[r:HAS_INTERVENTION]->(t:Intervention) 
where toLower(t.name) CONTAINS "remdesivir"
return distinct c

Case query 2

Retrieve all COVID-19 clinical trials that target “remdesivir” as the intervention but exclude pregnant women [OMOP ID: 4299535] from participating

match (c:Clinicaltrial)-[r:HAS_INTERVENTION]->(t:Intervention) 
where toLower(t.name) CONTAINS "remdesivir"
AND EXISTS {match (c)-[ic:EXCLUDE_CONDITION]->(condi:Condition) 
where condi.id = "4299535"}
return distinct c

OR

match (c:Clinicaltrial)-[r:HAS_INTERVENTION]->(t:Intervention)
where toLower(t.name) CONTAINS "remdesivir"
AND EXISTS {match (c)-[ic:EXCLUDE_CONDITION]->(condi:Condition)
where condi.name = "Pregnant"}
return distinct c

Case query 3

Retrieve all COVID-19 clinical trials that target “hydroxychloroquine” as the intervention and allow patients with shortness of breath [OMOP ID: 312437] to participate

match (c:Clinicaltrial)-[r:HAS_INTERVENTION]->(t:Intervention) 
where toLower(t.name) CONTAINS "hydroxychloroquine"
AND EXISTS {match (c)-[ic:INCLUDE_CONDITION]->(condi:Condition) 
where condi.id = "312437"}
return distinct c

OR

match (c:Clinicaltrial)-[r:HAS_INTERVENTION]->(t:Intervention)
where toLower(t.name) CONTAINS "hydroxychloroquine"
AND EXISTS {match (c)-[ic:INCLUDE_CONDITION]->(condi:Condition)
where condi.name = "Dyspnea"}
return distinct c

Case query 4

Retrieve all COVID-19 clinical trials in the United States that target “hydroxychloroquine” as the intervention and allow patients with diabetes to participate

match (c:Clinicaltrial)-[r:HAS_INTERVENTION]->(t:Intervention)
where toLower(t.name) CONTAINS "hydroxychloroquine"
AND EXISTS {match (c)-[rloc:HAS_LOCATION]->(loc:Location)
where toLower(loc.name) = "united states"}
AND EXISTS {match (c)-[ic:INCLUDE_CONDITION]->(condi:Condition)
where toLower(condi.name) CONTAINS "diabetes"}
return distinct c

We also provide eligibility criteria terms mapping file for the reference. Users can use this file ec_terms_mapping.tsv to look up clinical concepts and their OMOP IDs.

Please contact Jingcheng Du: [email protected], if you have any questions