Datasets:

everycure
/

kg-nodes

license: other
license_name: custom
license_link: LICENSE
dataset_info:
  features:
    - name: id
      dtype: string
    - name: name
      dtype: string
    - name: category
      dtype: string
    - name: description
      dtype: string
    - name: equivalent_identifiers
      list: string
    - name: all_categories
      list: string
    - name: publications
      list: string
    - name: labels
      list: string
    - name: international_resource_identifier
      dtype: string
    - name: upstream_data_source
      list: string
  splits:
    - name: train
      num_bytes: 5344685868
      num_examples: 9229096
  download_size: 1246629480
  dataset_size: 5344685868
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/nodes/train-*
task_categories:
  - graph-ml
language:
  - en
tags:
  - biology
  - medical
pretty_name: Every Cure Integrated Knowledge Graph

Dataset Card for Every Cure Integrated Knowledge Graph (Nodes)

Dataset Description

Homepage: https://docs.dev.everycure.org/releases/public_data_releases/
Paper: NA
Point of Contact: https://github.com/everycure-org/matrix/issues

Dataset Summary

The Every Cure KG is currently (as of February 2026) essentially an integrated, simplified and filtered merged KG comprising ROBOKOP and RTX-KG2. The "Nodes" datasets contains the records for all nodes in the graph, including metadata.

See edges dataset for the corresponding set of edges.

Source Data

Licensing Information

See LICENSE for details.