A prototype-augmented graph representation learning framework for identifying brain disorder-associated genes and facilitating drug repurposing
Clinical Snapshot
PICO Framework
| P — Population | Genetic loci and multi-omics data from patients with neuropsychiatric disorders (2 conditions) and neurodegenerative/neurological diseases (3 conditions), with experimental validation in primary neuron models |
| I — Intervention | Multi-omics Graph Transformer Network (MOGT) - a semi-supervised graph neural network for disease gene prediction |
| C — Comparator | Current approaches for disease gene prediction (specific methods not detailed in abstract) |
| O — Outcomes | Disease-associated gene prediction accuracy, identification of high-risk genes for Parkinson's disease, drug repurposing candidates, experimental validation of UK-356618 effects on PD-associated gene expression and cellular phenotypes |
Bottom Line
This computational study presents MOGT, a novel graph neural network approach for identifying disease-associated genes in brain disorders and facilitating drug repurposing. While the methodology appears innovative and shows promise for linking GWAS findings to disease mechanisms, the abstract lacks crucial details about performance metrics, statistical validation, and generalizability. The experimental validation of UK-356618 in primary neurons provides encouraging proof-of-concept, but broader clinical validation is needed. For Australian clinicians, this represents an emerging technology that may eventually inform precision medicine approaches for neurological and psychiatric conditions, though immediate clinical application is limited. The drug repurposing aspect could be particularly relevant given PBS considerations for cost-effective treatments. However, clinicians should await peer-reviewed publication of detailed methodology and performance data before considering clinical implications. The approach may eventually contribute to personalized treatment strategies for brain disorders, but requires substantial further validation in diverse populations and clinical settings.
Key Findings
P Value: Not reported
Effect Size: Not specified in abstract
Primary Outcome: MOGT outperformed current approaches in disease gene prediction across five brain disorders
Nnt Or Sensitivity: Sensitivity and specificity metrics for gene prediction not provided; 10 drug candidates identified for Parkinson's disease
Confidence Interval: Not reported
Clinical Application
Requires significant computational infrastructure and multi-omics data integration capabilities, limiting immediate clinical implementation Potential relevance to Australian precision medicine initiatives and TGA drug repurposing pathways, though validation in Australian populations would be required. Could inform PBS decision-making for repurposed medications if clinical trials demonstrate efficacy Patients with neuropsychiatric disorders and neurodegenerative diseases, particularly those with genetic risk factors identified through GWAS
Abstract
Many genetic loci were identified as associated with neuropsychiatric disorders and neurodegenerative disorders by Genome-wide association studies (GWAS). How these loci impact these diseases is unclear. Advances in deep-learning approaches and multi-omics data have the potential to link GWAS findings with disease mechanisms. Here, we proposed the Multi-omics Graph Transformer Network (MOGT), a semi-supervised graph neural network that leverages graph representation learning to model biological networks derived from multi-omics data to predict disease-associated genes. MOGT outperforms the current approaches in disease gene prediction for two psychiatric disorders and three neurodegenerative/neurological diseases. High-risk genes (HRGs) for Parkinson's disease (PD) predicted by MOGT were used to drug discovery by integrating with the CMAP database. Finally, 10 drugs were identified as potential candidates. Among them, the effect of drug UK-356618 was experimentally verified in a primary neuron model, showing that UK-356618 reversed the abnormal expression of PD-associated genes and improved the cell-level phenotypes of PD. Together, these results indicate that MOGT can be used to identify HRGs for brain disorders, and these predicted HRGs provide high-level insights into the mechanisms and treatments of brain disorders.
References
- 1.Li, J., Li, Y., Lin, S., Rao, J., & Zhao, H. (2026). A prototype-augmented graph representation learning framework for identifying brain disorder-associated genes and facilitating drug repurposing. PLoS Computational Biology. https://doi.org/10.21105/joss.03167
Related Research
Sleep medicine reviews
Machine and deep learning in REM sleep behavior disorder: a scoping review and analysis of reporting quality
3 Aug 2026
Neurological research
Federated deep learning model for epilepsy seizure detection using electroencephalogram (EEG) signal
2 Aug 2026
Journal of psychiatric research
Systematic review of machine learning and deep learning models for EEG-based detection of depression.
2 Aug 2026
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service