GenoDisc: an open-source functional genomics pipeline for drug repurposing and biomarker discovery applied to amyotrophic lateral sclerosis

Background: Genome-wide association studies identify numerous genetic risk loci, yet translating these statistical associations into actionable therapeutic targets and prognostic biomarkers remains a profound translational bottleneck. To bridge this gap, we developed the GenoDisc (Genomic Discovery) software pipeline, an open-source workflow automating complex post-GWAS functional analyses. Scheduled for public release next month, GenoDisc enables researchers to seamlessly conduct target identification and drug repurposing. This study introduces the pipeline’s architecture and demonstrates its clinical utility by applying it to genome-wide data for amyotrophic lateral sclerosis, a fatal neurodegenerative disease requiring urgent prognostic and therapeutic innovation.

Methods: The GenoDisc pipeline performs automated quality control, harmonisation, and integration of GWAS summary statistics with external functional genomic annotations, including tissue-specific expression and protein quantitative trait loci (eQTL/pQTL). We applied the pipeline to the latest European-ancestry amyotrophic lateral sclerosis GWAS (27,205 cases, 110,881 controls). To reduce methodological complexity, GenoDisc’s automated modules executed single-variant fine-mapping alongside transcriptome-wide and proteome-wide association studies (TWAS/PWAS). Furthermore, drug repurposing modules evaluated drug-gene interactions to identify disease-reversing compounds using reference databases like the Connectivity Map. Finally, the pipeline’s TWAS outputs were utilised to calculate predictive polytranscriptomic scores (PTS). This was achieved by aggregating observed whole-blood expression values weighted by their respective TWAS effect sizes. These scores were evaluated using linear regression against clinical characteristics in two independent whole-blood target cohorts.

Results: GenoDisc successfully distilled complex genome-wide data, packaging all outputs into an accessible interactive application. The gene-finding modules identified 118 high-confidence genes associated with the disease, capturing novel and known mechanistic drivers such as TMEM175 and SCFD1. Downstream drug repurposing analyses translated these findings into specific pharmacological insights, identifying six individually enriched compounds, including the nonsteroidal anti-inflammatory drug suxibuzone. Directional gene-set enrichment predicted that signatures linked to selective calcium channel blockers may reduce disease risk, while those associated with antiepileptics may exacerbate it. Finally, the calculated PTS proved to be a highly significant predictor of case-control status (liability R² = 5.1%, p = 3.2e-27). Within diagnosed patient cohorts, higher scores were significantly associated with adverse clinical features, including a higher likelihood of spinal site of onset, a later age at onset, and decreased overall survival.

Conclusions: The upcoming public release of GenoDisc provides the neuropsychopharmacology community with a scalable computational platform capable of translating raw GWAS data into testable pharmacological applications. By applying the pipeline to amyotrophic lateral sclerosis, we highlighted novel therapeutic avenues and demonstrated the viability of blood-based transcriptomic risk scores as non-invasive biomarkers. While constrained by the availability of diverse-ancestry reference data, this software significantly lowers technical barriers to advanced target discovery. During this presentation, we will demonstrate the pipeline’s interactive capabilities, providing the audience with a complete framework to progress from genetic discovery to actionable therapeutic targets and prognostic biomarkers.

References

1. Pain, O., Jones, A., Al Khleifat, A., Agarwal, D., Hramyka, D., Karoui, H., Kubica, J., Llewellyn, D.J., Ranson, J.M., Yao, Z. and Iacoangeli, A., 2024. Harnessing transcriptomic signals for amyotrophic lateral sclerosis to identify novel drugs and enhance risk prediction. Heliyon, 10(15).