Every Cure Knowledge Graph: A Unified Biomedical Knowledge Graph for Drug Repurposing
Identifying causal connections between existing drugs and mechanistic profiles of diseases is a foundational step for effective drug repurposing. Although knowledge graphs (KGs) are highly suited for consolidating biomedical databases and tracking these connections, a single biomedical KG is constrained by its ingestion pipeline and knowledge sources. While different biomedical KGs could be…
Curing diseases through repurposing existing drugs is a crucial step, yet knowledge graphs (KGs) often struggle to consolidate this information due to their disparate ingestion pipelines and limited sources. While combining these KGs could yield a more comprehensive resource, interoperability and provenance issues often impede such efforts.
To tackle these challenges, we introduce EC-KG, a biomedical KG tailored for computational drug repurposing, built using the Biolink Model and compatible with interoperable, provenance-focused approaches.
EC-KG unifies RTX-KG2, ROBOKOP, and PrimeKG at the network level, amalgamating over 7 million nodes and 81 million edges from 95 primary data sources. This integration yields enhanced coverage of core biomedical entities like drugs, targets, and diseases relevant to drug repurposing compared to individual source graphs. Furthermore, EC-KG captures intricate biomedical mechanisms within its topology, enabling the emergence of novel, mechanistically relevant pathways that are disconnected in the original constituent networks.
We demonstrate EC-KG's utility in method development, benchmarking, and predictive drug repurposing applications. Notably, it has been applied to identify Botulinum Toxin A as a potential treatment for Major Depressive Disorder and to validate the repurposing of Lenalidomide and Dexamethasone for a subgroup of patients with Rosai-Dorfman Disease. EC-KG's success underscores its potential as a valuable resource for advancing drug repurposing research.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
