Integrative drug discovery for Parkinson disease with DIMER
- 1Emory University
Abstract
Translating Parkinson disease (PD) genetics into therapeutic candidates requires approaches that account for interconnected disease mechanisms and their cellular context. Here we present DIMER (Direct and Indirect Module Evaluation for Repurposing and screening), a framework that connects network-based drug repurposing and target-directed virtual screening through shared disease-associated gene sets. Applying DIMER to PD, we integrated genetic risk and mechanism-associated gene sets with drug-target interactions, disease transcriptomes, and single nucleus expression profiles. Network analysis identified 423 repurposing candidates, with 71% of qualifying drug-gene set associations arising through targets outside the corresponding disease gene set. Transcriptional reversal analysis and cell type-resolved networks further distinguished candidates by their predicted effects on disease-associated expression and their cellular context, while target-set separation nominated potentially complementary drug combinations. For target-directed virtual screening, DIMER adapted a pretrained compound-protein interaction model to screening library chemistry and incorporated known-ligand similarity into compound ranking. Library adaptation improved compound-target pair classification performance, whereas ligand-informed ranking increased early enrichment of known active compounds. Consensus screening of 1.82 million compounds against seven PD-associated proteins yielded a shortlist of 22 compounds supported by at least five ranking runs. In a complementary experimental evaluation, four repurposing candidates (polydatin, fedratinib, sorafenib and venlafaxine) attenuated the loss of dopamine neuron marker signals in human midbrain organoids exposed to 6-hydroxydopamine. Together, these findings establish DIMER as a framework for translating disease-associated genes into pharmacological hypotheses and prioritizing candidates for experimental testing.
Related
- Integrative Transcriptomic Meta-Analysis Reveals SIRT1-SNCA Correlation in Parkinson's Disease
- Mapping human genetic associations to drug targets in valvular heart disease: an integrative evidence synthesis
- Integrating Multisource Target Evidence with Machine-Learning Potency Prediction for Alzheimer's Disease Drug Discovery
- Integrative Single-Cell Transcriptomics and Explainable AI for Disease Mechanism and Biomarker Discovery in Triple-Negative Breast Cancer
- Integrative omics reveals candidate drug-repurposing targets in Tritrichomonas foetus
BibTeX
@misc{guo2026integrative,
title = {{Integrative drug discovery for Parkinson disease with DIMER}},
author = {S. Guo and X. Yang and Y. Dong and L. Yu and A. Sangjizhuoma and K. L. Sie and K. Zhu and Y. Zhang and X. Zhao and E. Kerman and R. Li and J. He},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.28.762666},
url = {https://www.biorxiv.org/content/10.64898/2026.10.28.762666v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Integrative drug discovery for Parkinson disease with DIMER
AU - S. Guo
AU - X. Yang
AU - Y. Dong
AU - L. Yu
AU - A. Sangjizhuoma
AU - K. L. Sie
AU - K. Zhu
AU - Y. Zhang
AU - X. Zhao
AU - E. Kerman
AU - R. Li
AU - J. He
PY - 2026
DA - 2026/10/28
PB - bioRxiv
DO - 10.64898/2026.10.28.762666
UR - https://www.biorxiv.org/content/10.64898/2026.10.28.762666v1
AB - Translating Parkinson disease (PD) genetics into therapeutic candidates requires approaches that account for interconnected disease mechanisms and their cellular context. Here we present DIMER (Direct and Indirect Module Evaluation for Repurposing and screening), a framework that connects network-based drug repurposing and target-directed virtual screening through shared disease-associated gene sets. Applying DIMER to PD, we integrated genetic risk and mechanism-associated gene sets with drug-target interactions, disease transcriptomes, and single nucleus expression profiles. Network analysis identified 423 repurposing candidates, with 71% of qualifying drug-gene set associations arising through targets outside the corresponding disease gene set. Transcriptional reversal analysis and cell type-resolved networks further distinguished candidates by their predicted effects on disease-associated expression and their cellular context, while target-set separation nominated potentially complementary drug combinations. For target-directed virtual screening, DIMER adapted a pretrained compound-protein interaction model to screening library chemistry and incorporated known-ligand similarity into compound ranking. Library adaptation improved compound-target pair classification performance, whereas ligand-informed ranking increased early enrichment of known active compounds. Consensus screening of 1.82 million compounds against seven PD-associated proteins yielded a shortlist of 22 compounds supported by at least five ranking runs. In a complementary experimental evaluation, four repurposing candidates (polydatin, fedratinib, sorafenib and venlafaxine) attenuated the loss of dopamine neuron marker signals in human midbrain organoids exposed to 6-hydroxydopamine. Together, these findings establish DIMER as a framework for translating disease-associated genes into pharmacological hypotheses and prioritizing candidates for experimental testing.
ER -