Overview

• Researchers at Harvard Medical School reported PDGrapher in Nature Biomedical Engineering and made the tool freely available for research use.
• The causally inspired graph neural network integrates protein–protein interactions and gene regulatory data to prioritize targets that move diseased expression profiles toward healthy states.
• In benchmarks across genetic and chemical perturbation datasets spanning 11 cancer types, the model surpassed competing methods in ranking accuracy and delivered results up to 25 times faster.
• PDGrapher recovered withheld, clinically supported targets such as KDR in non‑small cell lung cancer and prioritized TOP2A in certain tumors, while tracing mechanisms for drugs like vorinostat and sorafenib.
• The team is extending the platform to neurodegenerative diseases including Parkinson’s and Alzheimer’s with a collaboration on X‑linked dystonia‑parkinsonism at Massachusetts General Hospital, with patient‑specific applications contingent on further validation.