Explore the BrainHBP Knowledge Graph Draft!


HBP believes in the promise of Open Science.

We built the Knowledge Graph to help you find and share the data you need to make your next discovery. And we built it to connect you to the software and hardware tools which will help you analyse the data you have and the data you find.

The HBP Knowledge Graph supports rich terminologies, ontologies and controlled vocabularies. The system is built by design to support iterative elaborations of common standards and supports these by probabilistic suggestion and review systems.

The HBP Knowledge Graph is a multi-modal metadata store which brings together information from different areas of the Human Brain Project as well as from external partners. At the core of the HBP Knowledge Graph, a graph database tracks the linkage between experimental data and neuroscientific data science supporting more extensive data reuse and complex computational research than would be possible otherwise.

If you don't find what you're looking for contact us at: kg-team@humanbrainproject.eu

Have fun!

Explore the knowledge graph

Knowledge Graph Search

Search for data and data connections with an easy-to-use facet search interface.

Browse the graph!
Knowledge Graph Statistics

How many entities is in the Knowledge Graph? How is it connected and what are the internal datastructures? All these questions can be answered by KG statistics.

Check out the structure!
Knowledge Graph Editor

Edit the metadata inside the Knowledge Graph and create new instances supported by input aids and built-in ontologies.

Change the metadata!

The platform

The HBP Knowledge Graph is a combination of multiple system components built to address the various needs of the data providers and consumers.

How to get your data into the graph

You have data you would like to share? Perfect! The HBP Knowledge Graph has the concept of private spaces and is – due to it's nature as a graph – very permissive in terms of the structure of data that can be added to it!

All you need to start getting your data in is:

The data format of choice for the knowledge graph is JSON-LD which allows to enrich the well-known JSON structure with semantic information.

The data entry into the HBP Knowledge Graph is backed by the Blue Brain Nexus. This means that your data can be entered and edited either through the API, the python convenience library Pyxus or the HBP Knowledge Graph Editor. We are aware that your raw-data is most likely not stored in the JSON-LD format - we do have experiences in transformation of CSV, XLS, JSON as well as the full integration of external APIs. Ask us!

Your data can theoretically be freely structured - but it only will profit from linkage to other information and therefore from the additional enrichment if it is aligned with / following some conventions of other data.

Nexus – and therefore also the HBP Knowledge Graph – provides the possibility to define validation schemas in SHACL to ensure the uploaded data is consistent with the defined structure. Since we follow an open-world approach, the strictness of those schemas can be defined freely. We suggest to keep them very permissive in the beginning and let them become more strict over time when the used data structures gain in maturity.
Depending on your domain it may also make sense to reuse or extends some common models and naming conventions such as schema.org or the INCF neuroshapes.

How to access data

Although some basic access to the data is available through the Nexus API / Pyxus, we recommend to make use of the HBP KG Query API.

By using a simple specification language, you can declare WHAT (the scope) of the graph you would like to query from the graph, HOW the response shall be formatted (JSON, CSV, XML, ...) and WHERE the resulting structures should be provided (as a JSON file, indexed in Elasticsearch, Apache Solr, ...). You define your own API on the graph or restrict an already existing one by applying your custom template. This has multiple advantages:

Create your own HBP Knowledge Graph API

See how you can easily build your own Knowledge Graph API optimized for your specific use-case.

Design your API!
Upload your data with Python

Find example use of the data upload from within a Jupyter Notebook

Upload your data!
Read your data with Python

Find example use how data can be read within a Jupyter Notebook by the use of Pyxus

Read your data!

HBP specifics / conventions

Private spaces

The primary structure of the HBP Knowledge Graph is split into multiple private spaces. Private spaces are always self-contained, their data comes from a single origin (single data master) and they are composed of 3 spaces following naming conventions:

These spaces should be seen as one concerning the user perspective. They do differ in their definition of access rights (see below). Please note, that the "{spacename}reconciled" is read only even for administrators, since it is populated automatically.

Autopopulated spaces

Just as the "{spacename}"reconciled space, there are other spaces which are automatically populated based on the information coming from the different private spaces. By taking into account different criterias, new spaces are built (e.g. based on a API specification) by scripts to simplify access, stage reduced / extended data to spaces accessible by specific usergroups as well as to support easy data transformation.

Access rights

Access rights are defined on the level of "private spaces" - we have three user groups:


Any questions? Contact us: kg-team@humanbrainproject.eu