Biomedical KB-HD is distributed through: https://my.qiagendigitalinsights.com/bbp/
For access, please contact BKB support: ts-bioinformatics@qiagen.com (IP address is needed for secured access)
Biomedical KB-HD includes:
Depending on your license, it may also include:
BKB-HD API is a QIAGEN-hosted SQL query service that provides access to BKB-HD through Python or R BKB-HD packages or directly through REST API.
kb = bkb.BKb.from_api()kb = bkb.BKb.from_api(device_code_flow=False)kb = bkb.BKb.from_api(username='your_username', password='your_password')bkb <- bkb::from_api()bkb <- bkb::from_api(device_code_flow=FALSE)bkb <- bkb::from_api(username='your_username', password='your_password')More resources:
In general, memory and system requirements depends on the type of analysis performed. For standard queries using SQL or the R/Python packages, 16GB is sufficient, but some types of analysis (ex. using a full-graph, in-memory NetworkX export) may require more than 32GB of memory.
Python:
Python 3.9 or later is required. The system must have an SQLite version 3.8.3 or later.
The Python package provides wrapper functionality for the Biomedical KB-HD. The Python package can be downloaded to a local folder and installed as: pip3 install bkb-2025.1-py3-none-any.whl
To start using the package, you will need to connect to either BKB-HD API, an SQLite database or SQL DBMS (such as PostgreSQL).
R package:
R 4.1.0 or later is required. Additionally, the following packages must be present: 'dplyr', 'dbplyr', 'DBI', 'R6', 'assertthat', 'RSQLite', 'tibble', 'igraph', 'websocket', 'future.apply', 'readr', 'httr', 'fs', 'shiny', 'htmlwidgets' and 'uuid'
The 'bkb' R package provides convenience functions for working with the Biomedical KB-HD. To get started with the bkb R package, download the 'bkb_2025.1.tar.gz' file to your local machine.
Neo4j:
For the Neo4j export, Neo4j version 5.5 is the recommended version. We do not test earlier versions but expect them to be compatible (the Python graph export scripts support both Neo4j version 4 and version 5). The decompressed Neo4j files will take up ~5 GB, and after installation ~7 GB will be used.
SQLite:
The SQLite file is ~10GB compressed, and will decompress to ~50GB, which means at least 60 GB of disk space must be available for the decompression to succeed. SQLite for Causal Reasoning: For the causal reasoning functionality and notebooks, SQLite version 3.35.3 or later is required.
PostgreSQL:
The PostgreSQL import script has been tested on PostgreSQL 14.1 on Linux. It will take up at least 50 GB of disk space.
Our support team: ts-bioinformatics@qiagen.com
Contact them for scientific and technical support, including license activation questions and BKB access.