The code in this repo can reproduce the second Kibi ChEMBL demo. This are the steps to reproduce the demo.

1. Start up a kibi instance. E.g. using docker: docker run -d -p 5606:5606 -p 9201:9220 --net=chembl --name chembl-kibi sirensolutions/kibi-community-standard:4.6.4

2. pip install -r requirements.txt

3. run python import.py -es http://localhost:9201 using the port your elasticsearch instance is exposed to and wait for it to complete this will:

   • Download and extract the chembl dumps in SQLite format
   • Query the database and generate the input json file in the import directory
   • Load the json in elasticsearch with the proper mapping
   • Create the necessary index-pattern in kibi

4. explore the data at your kibi instance: http://localhost:5606

5. To persist the state of kibi you can run dump_kibi.sh. This requires elasticdump to be installed. To save the data and the kibi configuartion in a compressed file you can run dump_all.sh.