Medical Knowledge Graph Deployment and Usage

Description

This repository is for the Medical Knowledge Graph (MKG) deployment and usage.

For more details see our paper: Medical Knowledge Graph to Promote Rational Drug Use: Model Development and Performance Evaluation

If you use this MKG, please cite:

@article{liao2022medical,
  title={Medical Knowledge Graph to Promote Rational Drug Use: Model Development and Performance Evaluation},
  author={Liao, Xiong and Liao, Meng and Guo, Andi and Luo, Xinran and Li, Ziwei and Chen, Weiyuan and Li, Tianrui and Du, Shengdong and Jia, Zhen},
  journal={Human-Centric Intelligent Systems},
  volume={2},
  number={1},
  pages={1--13},
  year={2022},
  publisher={Springer}
}

MKG Data

• The MKG dump for deployment is at Google Drive.
• The total data size is around 1 GB.

Linux installation(Ubuntu)

This is an operational manual for deploying Neo4j on Ubuntu. If you want to learn more information or deploy it on other systems, you can refer to here.

Java prerequisites

Neo4j 5 requires the Java 17 runtime.

1.List all your installed versions of Java:

update-java-alternatives --list

1.1 If Java 17 is not included in the results, execute the following command to install OpenJDK 17:

sudo apt install openjdk-17-jre

1.2 Set it as the default by replacing <java17name> with its name:

sudo update-java-alternatives --jre --set <java17name>

2.Identify your Java 17 version:

java -version

Installation

1.To use the repository for generally available versions of Neo4j, run:

wget -O - https://debian.neo4j.com/neotechnology.gpg.key | sudo gpg --dearmor -o /etc/apt/keyrings/neotechnology.gpg

echo 'deb [signed-by=/etc/apt/keyrings/neotechnology.gpg] https://debian.neo4j.com stable latest' | sudo tee -a /etc/apt/sources.list.d/neo4j.list

sudo apt-get update

2.Install neo4j community edition:

sudo apt-get install neo4j=1:5.26.1

3.Starting the service automatically on system start:

sudo systemctl enable neo4j

4.Then you can run the following command to controlling the service:

sudo neo4j {start|stop|restart}

Start Neo4j

1.Modify the configuration file, located at /etc/neo4j/neo4j.conf:

initial.dbms.default_database=graph

2.Restart neo4j service:

sudo neo4j restart

3.Open http://localhost:7474 in your web browser.

4.Connect using the username neo4j with your password or the default password neo4j. You will then be prompted to change the password. The default database is graph which you set in step 1.

Import database

1.Stop neo4j service:

sudo neo4j stop

2.Create a dump folder at $NEO4J_HOME/ :

cd <NEO4J_HOME>
sudo mkdir dump

3.Place graph.dump inside it, run:

neo4j-admin database load --from-path=dump graph --overwrite-destination=true

4.Start neo4j service, then you can use this database:

sudo neo4j start

APOC

• APOC (Awesome Procedures on Cypher) is an add-on library for Neo4j that provides hundreds of procedures and functions adding a lot of useful functionality. You can find more information in here.

• APOC can be installed by moving the APOC jar file from the $NEO4J_HOME/labs directory to the $NEO4J_HOME/plugins directory and restarting Neo4j.

MKG Usage Examples

We provide a script drug_review_system.py that shows how to connect to and retrieve the MKG to verify a patient's medication safety based on rules and use cases.

• Requirement: neo4j package.

  pip install neo4j
  

• query_database function: call the run method on the graph object to execute a query in the MKG. The query statement is written in Cypher, a declarative graph query language used by Neo4j.

• drug_interactions function: check for potential interactions between a set of drugs a patient might take by using the DDI knowledge in the MKG.

• allergy_review function: check for any allergens in the drugs a user takes.

• adverse_reaction_review function: check for potential adverse reactions that a user might experience based on the drugs they are taking and a disease that they are suffering.

• duplicate_drug_review function: check for potential duplicate or overlapping ingredients between different medications a user might be taking.

• contraindication_review function: check for contraindications between a list of drugs that a patient is taking and a specified disease.

• age_review function: review whether the drugs a patient takes suit their age.

• special_population_review: review whether the drugs a patient takes suit special populations, such as pregnant women, children, elderly people, or those with specific medical conditions.

• method_review function: review whether the methods for the drugs a patient is taking are correct.

Feedback

Any feedback is welcome! If you have any issues downloading or deploying the MKG, please do not hesitate to contact us via mail: zjia@swjtu.edu.cn.

License

The MKG is licensed under MIT license.