SPIDER-KG

This dataset contains the original static release of SPIDER-KG, including merged biomedical entities and relations, source provenance, standardized relation mappings, historical literature evidence, condition extraction results, conflict classifications, and identifier mappings. It excludes subsequent daily updates. The base graph contains 3,748,007 entities and 12,752,997 resolved directed native entity pairs. Historical literature evidence is stored in a separate SQLite database for a selected subgraph. Condition extraction and conflict classification results include 347,916 and 89,937 records, respectively. See the README and data dictionary for scope and known limitations. Code: https://github.com/MAGIC-AI4Med/SPIDERWebsite: https://spiderkg.internal.onexai.com.cn/