Repurposing Existing Drugs for Neurodegenerative Diseases Using a Systematic Deep Learning Analysis of Multi-Omic Data

Shefali Sharma ¹, Dr. T. Vengatesh ^1*, D. Mythili ², M. Parimala ³, Dr. P. Muthu Pandian ⁴, Dr. M. Geethalakshmi ⁵, Saint Jesudoss S ⁶, K N V Ramya Devi ⁷, P. Boomi ⁸, Viswanathan Ramasamy ⁹

¹Assistant Professor, New Horizon College of Engineering, Bengaluru, Karnataka, India.
Email: shefalisharma888@gmail.com

^1*Assistant Professor, Department of Computer Science, Govt. Arts & Science College, Theni, Affiliated to Madurai Kamaraj University, Madurai, Tamilnadu, India.
Email: venkibiotinix@gmail.com (Corresponding Author)

²Assistant Professor, Department of Mathematics, Erode Sengunthar Engineering College, Perundurai - 638057, Tamilnadu, India.
Email: mythimkp@gmail.com

³Associate Professor, Department of Mathematics, Bannari Amman Institute of Technology, Sathyamangalam, Tamilnadu, India.
Email: rishwanthpari@gmail.com

⁴Assistant Professor, Department of Chemistry, Saveetha Engineering College, Thandalam, Chennai - 602105, Tamilnadu, India.
Email: muthu.pandian500@gmail.com

⁵Associate Professor, Department of Mathematics, KCG College of Technology, Karapakkam, Chennai, Tamilnadu, India.
Email: geetharamon@gmail.com

⁶Assistant Professor, Department of CSE, Rajiv Gandhi College of Engineering and Technology, Puducherry, India.
Email: saint.2k5@gmail.com

⁷Assistant Professor, Department of IT, S.R.K.R. Engineering College, Bhimavaram - 534204, Andhra Pradesh, India.
Email: kotla.ramya@gmail.com

⁸Assistant Professor, Department of Mathematics, V.S.B. Engineering College (Autonomous), Karur, Tamilnadu, India.
Email: rpboomi@gmail.com

⁹Professor, Department of CSE, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Andhra Pradesh, India.
Email: rvnathan06@gmail.com

*Corresponding Author: Dr. T. Vengatesh, Assistant Professor, Department of Computer Science, Govt. Arts & Science College, Theni, Affiliated to Madurai Kamaraj University, Madurai, Tamilnadu, India. Email: venkibiotinix@gmail.com

---

ABSTRACT

Neurodegenerative diseases (NDDs), including Alzheimer's (AD) and Parkinson's (PD), represent a growing global health burden with limited therapeutic options. Drug development is notoriously costly, time-consuming, and high-risk. Computational drug repurposing offers a promising strategy to identify novel therapeutic uses for existing approved drugs. This study proposes a systematic deep learning framework that integrates heterogeneous multi-omic data including genomics, transcriptomics, proteomics, and epigenomics to predict novel drug-disease associations for NDDs.

We construct a multi-layered biological network incorporating disease-specific perturbations from patient-derived omics data and drug-induced signatures from connectivity databases (e.g., LINCS). A graph neural network (GNN) model is trained to learn latent representations of drugs and diseases, capturing complex, non-linear relationships within and between omic layers. Our model identifies several high-probability repurposing candidates, such as dasatinib (an oncology drug) for AD and bromhexine (a mucolytic) for PD, based on predicted reversal of disease-associated gene expression patterns.

Experimental validation through in silico pathway analysis and literature mining supports the biological plausibility of these predictions. This work demonstrates that systematic integration of multi-omic data using deep learning can accelerate the discovery of viable, mechanistically supported repurposing opportunities for neurodegenerative diseases.

Keywords: Drug repurposing, neurodegenerative diseases, deep learning, multi-omic data, Alzheimer's disease, Parkinson's disease, graph neural networks, bioinformatics.

How to cite this article: Sharma S, Vengatesh T, Mythili D, Parimala M, Pandian PM, Geethalakshmi M, Jesudoss SS, Devi KNVR, Boomi P, Ramasamy V. Repurposing existing drugs for neurodegenerative diseases using a systematic deep learning analysis of multi-omic data. Int J Drug Deliv Technol. 2026;16(3s): 909-918; DOI: 10.25258/ijddt.16.3s.112

Source of support: Nil.

Conflict of interest: None