The Bio2KG Registry is a repository of dataset descriptions, including preferred CURIE prefixes, base URIs, identifier regex patterns and HTML resolvers for biomedical datasets.

The registry is constructed from a manually curated spreadsheet.

1. Extract data from the Life Science Registry spreadsheet on Google docs
2. Load to ElasticSearch (deployed with the docker-compose.yml file)

Access the Bio2KG registry web application and API endpoints:

• Search website: https://registry.bio2kg.org

• GraphQL API: https://registry.bio2kg.org/graphql

• ElasticSearch API: https://elastic.registry.bio2kg.org/_search

Search with cURL:

curl -XGET --header 'Content-Type: application/json' https://elastic.registry.bio2kg.org/prefixes/_search -d '{
      "query" : {
        "match" : { "Preferred Prefix": "bio" }
    }
}'

The process to prepare the ElasticSearch index for the Life Science Registry runs in a docker container defined in the etl folder.

To update a running docker-compose stack:

docker-compose run update-pipeline

1. Prepare the permission for the shared volume to keep ElasticSearch data persistent:

sudo mkdir -p /data/bio2kg/registry/elasticsearch
sudo chmod g+rwx -R /data/bio2kg/registry/elasticsearch
sudo chgrp 1000 -R /data/bio2kg/registry/elasticsearch
sudo chown 1000 -R /data/bio2kg/registry/elasticsearch

2. Start the stack with docker-compose:

docker-compose up -d

• Search UI frontend on http://localhost:3000
• GraphQL server on http://localhost:4000/graphql
• ElasticSearch on http://localhost:9200

🔁 If you want to update the ElasticSearch endpoint data without stopping the stack, you can run this:

docker-compose run update-pipeline

The stack deploys:

• An ElasticSearch instance with a nginx proxy to allow anyone to access the /_search endpoint, but prevents editing, configuration defined in the elasticsearch folder on http://localhost:9200
• A NodeJS server using Searchkit and NextJS defined in the website folder on http://localhost:3000
  • Apollo GraphQL endpoint serving data from ElasticSearch on /graphql
  • A React website using SearchKit to search the data on the base URL (/)
  • 🚧 In development: a Sofa API to publish an OpenAPI endpoint based on the GraphQL endpoint
    • API on /api
    • Swagger UI on /apidocs

You can also start just the website using yarn without of docker, it will get the data from the ElasticSearch in production:

cd website
yarn
yarn dev

Start the stack with production config, using nginx-proxy to route the services:

docker-compose -f docker-compose.yml -f docker-compose.prod.yml up -d

Install the Linked Data Platform in the Virtuoso triplestore (running via docker-compose)

./prepare_virtuoso_ldp.sh

🔁 If you want to update the ElasticSearch endpoint data without stopping the stack, you can run this:

docker-compose run update-pipeline

To add a new field to the Bio2KG registry, check the following files:

1. In the etl folder python script: get the field from the spreadsheet, process it, and add the field value to ElasticSearch
2. In website/pages/api/graphql.tsx: add the field in the entry fields to register this field in the GraphQL server query
3. In website/components/index.tsx: add the field to the GraphQL query used by the UI to retrieve data.
4. In website/components/searchkit/Hits.tsx: add the field to the UI

⚠️ Make sure the id you are using for the field is the same everywhere! (it is case sensitive)