From Bench to Bedside:

AI-Enabled Drug Repurposing for Innovative Therapies in Complex Diseases

Karina Laxman Yadav¹*, Neha Desai², Anuradha Prajapati², Sachin Narkhede², Sailesh Luhar²

¹Smt. B.N.B. Swaminarayan Pharmacy College,

Gujarat Technological University, Salvav, Vapi, Gujarat, India, 396191.

²Department of Pharmaceutics, Smt. BNB Swaminarayan Pharmacy College,

Gujarat Technological University, Salvav, Vapi, Gujarat, India, 396191.

*Corresponding Author E-mail: yadav1212karan@gmail.com

ABSTRACT:

This comprehensive review article gives the information about the importance about AI in drug repurposing for complex diseases. The increasing prevalence of complicated diseases such as infectious diseases, neurological disorders, and cardiovascular problems calls for the development of novel therapeutic approaches. Conventional medication development is expensive and time-consuming; years of study and clinical trials are frequently needed. AI-enabled medication repurposing has surfaced as a viable method to accelerate the process in this regard. This methodology leverages artificial intelligence (AI) technology, including machine learning and natural language processing, to explore the possibilities of currently licensed drugs for novel therapeutic applications. The present level of AI-driven medication repurposing is examined in this review, with an emphasis on the benefits it offers over conventional techniques and how quickly it could lead to the discovery of new treatments. Key AI strategies, effective case studies, and the shift from preclinical to clinical research are covered. It also tackles difficulties including interpretability of models, data quality, and regulatory concerns. All things considered, AI-enabled drug repurposing is a revolutionary strategy for creating more efficient and reasonably priced treatments for complicated illnesses.

KEYWORDS: AI-enabled drug repurposing, complex diseases, precision medicine, machine learning, deep learning, network analysis.

INTRODUCTION:

Drug repurposing involves utilizing an existing drug or drug candidate for a new treatment or medical condition that it was not originally intended to address. This approach often emerges serendipitously, with the unintended side effects of a drug sometimes indicating its potential effectiveness for a different condition. Typically, these drugs have already established safety profiles in humans and have been tested for efficacy against diseases other than the one for which they were initially developed. This method allows for a more direct transition to preclinical and clinical trials, bypassing the lengthy drug development process and thereby reducing associated risks and costs.¹

The development of novel medicines for complicated diseases presents the pharmaceutical industry with unprecedented obstacles, with anticipated expenditures surpassing $2.5 billion and delays exceeding a decade. Due to its reliance on de novo design and synthesis, the traditional paradigm for drug development frequently results in high attrition rates and unmet medical requirements. On the other hand, finding new therapeutic uses for already-approved medications, or drug repurposing, presents a viable substitute.²The field of medication repurposing has undergone a radical change with the introduction of artificial intelligence (AI), which has made it possible to analyse enormous databases, spot unique patterns, and anticipate possible therapeutic uses. Repurposed indications for well-established medications have been successfully found by AI-enabled drug repurposing, exhibiting increased cost-effectiveness, safety, and efficacy.²

AI-driven drug repurposing identifies new treatment options for rare diseases using existing medications. By rapidly analyzing data, it reduces development time and costs while leveraging the safety profiles of approved drugs to minimize risks and address treatment gaps. AI integrates biomedical data to uncover drug-disease relationships, supports personalized medicine, and fosters collaboration among researchers. Overall, this approach enhances the efficiency and accessibility of therapies for rare diseases.

Figure 1. AI-based drug repurposing

AI-POWERED APPROACHES FOR DRUG REPURPOSING:

· Machine learning approach

· Deep learning approach

· Natural Language Processing

· Network analysis

· Graph-based methods

1. Machine learning approach:

To forecast new drug-disease connections, algorithms utilize extensive datasets of drug-target interactions, side effects, and disease pathways. AI innovation significantly enhances drug design through improved machine learning techniques and the aggregation of pharmacological data. Focusing on converting medical data into practical research, algorithms like Random Forest, Naive Bayes, and Support Vector Machines are commonly used. Increasingly, deep learning methods enhance feature extraction and generalization, employing both supervised techniques for predictive outcomes and unsupervised techniques for data clustering.^3,4

2. Deep learning approach:

An increasingly lucrative field in both business and research is deep learning, a crucial part of machine learning that extracts information from data using multi-layered neural networks. Particularly in drug discovery, methods like greedy layer-by-layer modeling aid in the generation of complicated models like DNNs, CNNs, RNNs, and autoencoders. By working independently on feature maps, pooling layers decrease spatial dimensions and computations, assisting in the reduction of overfitting and accelerating feature extraction.⁵

SUCCESSFUL APPLICATION OF AI-DRIVEN DRUG REPURPOSING:

1. Leveraging AI for Drug Repurposing in the Fight Against COVID-19⁶:

COVID-19 is a highly contagious disease that is caused by the SARS-CoV-2 coronavirus. It was first discovered in Wuhan, China, in December 2019. The disease spread quickly, resulting in a global pandemic. Common symptoms include fever, exhaustion, coughing, difficulty breathing, and loss of taste and smell. Drug repurposing involves using approved medications to treat new diseases like COVID-19, reducing development costs and time. In the big data era, AI and network medicine enhance the precise identification of therapeutic targets. This review provides recommendations for using AI in drug repurposing, highlighting its importance during urgent epidemics. Examples of repositioned drugs include hydroxychloroquine, remdesivir, ivermectin, and baricitinib. The study also addresses ongoing research into new COVID-19 treatments and recent advancements in the field.⁶

2.Leveraging AI for Drug Repurposing in Neurological Disorders:

About 43.8 million people worldwide suffer with Alzheimer's disease (AD), a chronic neurodegenerative disorder characterized by cognitive decline and neuronal loss. In order to find new therapeutic targets using GWAS-identified variations, this study proposes an artificial intelligence (AI) framework that integrates multi-omics data with protein-protein interaction networks. We identify AD risk genes (ARGs) by using a Bayesian method, and we rank repurposable medications according on how close these genes are to each other. We use in vitro investigations in human microglia cells and large-scale patient data to validate medication correlations with AD. Our results indicate pioglitazone as a possible treatment, highlighting the framework's ability to combine genetic discoveries with clinical applications, which may also improve the development of medicines for other complicated disorders.⁷

3. Leveraging AI for Drug Repurposing in Cardiovascular Disease:

Cardiovascular disease (CVD) remains a significant global health issue, causing high morbidity and mortality rates that strain healthcare systems. Traditional drug development methods are often costly and inefficient, making it challenging to find effective, individualized treatments. Artificial intelligence (AI) offers a transformative approach to drug discovery by analyzing large biological datasets to uncover patterns that humans might miss. This capability enhances treatment strategies, improves drug candidate design, and better predicts efficacy and toxicity, streamlining the search for new cardiovascular medications.⁸

4. Leveraging Artificial Intelligence for Drug Repurposing in Infectious Diseases:

Artificial intelligence (AI) is accelerating drug repurposing and finding new applications for current treatments, changing the battle against infectious diseases. This strategy integrates various biological data, such as protein structures and metabolic pathways, to uncover pathogen weaknesses, enabling faster and more economical therapy development.

To build prediction models, artificial intelligence (AI) computers will examine data from electronic medical records and Next Generation Sequencing (NGS). Instead of focusing only on specific proteins, sophisticated methods such as genome-scale metabolic modeling enable the targeting of important nodes within biological networks.AI can also be used to simulate the creation and improvement of drug candidates, which could result in safer, more effective medications. In order to fully realize AI's potential to improve global health outcomes and enable more focused therapies, cooperation and open data exchange are essential.⁹

Databases and tools for AI-driven drug repurposing:¹⁰

Table IV. I. Drug databases

Database	Description
ChEMBL	A carefully selected collection of bioactive compounds with characteristics similar to those of pharmaceuticals, combining chemical, bioactive, and genetic information to speed up the creation of novel, potent medications based on genetic data.
DrugBank	It integrates drug data such as chemical, pharmacological, and pharmaceutical information with drug target information, including sequence, structure, and pathways.
Pubchem	It offers a wide range of molecular data, such as patents, literature citations, biological activities, safety and toxicity information, chemical structure and physical properties, and so forth.

2. Disease databases:

Data Collection:

· Clinical Data

· Genomic Data

· Pharmacological Data

· Literature

Data Structuring:

· Disease Ontology

· Drug Information

· Interactions

AI and Machine Learning Integration:

· Predictive Modeling

· Natural Language Processing

· Network Analysis

Validation and Testing:

· In Silico Validation

· Experimental Validation

· Data Visualization

· Search Functionality

Collaboration and Updates:

· Open Access

· Regular Updates

3. AI-powered platforms

· Atomwise

· Insilico Medicine

· Benevolent AI

· Deep Genomics

· Recursion Pharmaceuticals

Future Pathways and Challenges Slow:

In the 2020s, artificial intelligence is set to make significant strides, especially in drug repurposing. This progress will blend data-driven and knowledge-driven methodologies, moving away from the limitations of opaque 'black box' systems to create more transparent and inclusive solutions for stakeholders. A key aspect of this advancement is leveraging the extensive knowledge contained in scientific literature. By analyzing the relationships among various entities, we can build a solid framework that improves our understanding of drug interactions and biological processes. High-resolution in silico physiological models, combined with real-time computer-assisted design, will help identify niche drug indications that align with specific target product profiles. Despite this potential, several methodological challenges persist. The lack of substantial positive data restricts machine learning techniques from effectively distinguishing between relevant and irrelevant entities. There is a pressing need for more efficient stratification and weighting algorithms to enhance the predictive accuracy of classifiers. Furthermore, developing automated explanation methods is essential for ensuring transparency in AI-driven approaches.^11,12

The implementation of FAIR (Findable, Accessible, Interoperable, Reusable) principles should facilitate drug repositioning rather than create barriers. Future datasets should include groups of known off-label drug uses, termed "Hit indications," while also accurately representing negative examples, or "Hard Failures," to improve the assessment of repositioning potential. Gaining insights into individual metabolomic traits will be crucial for identifying which patients are most likely to benefit from repurposed drugs. A biology-first approach will enrich our understanding of disease progression, potentially uncovering unexpected drug repositioning opportunities. The future of AI in drug repurposing may encompass more advanced versions of current methods, the introduction of innovative strategies, or a mix of both. Techniques that utilize extensive datasets and sophisticated representation systems—like chemical fingerprints integrated with network graph theory—are anticipated to evolve, aided by larger training datasets and enhanced stratification methods that optimize predictive power. By effectively managing knowledge, we can tap into the vast information available in scientific literature to propel drug discovery forward.^11,12

CONCLUSION:

Using AI to enable drug repurposing is a revolutionary step in the creation of novel treatments for difficult-to-treat illnesses. Utilizing currently available molecules and quickening the process of discovery may allow us to more effectively address unmet medical needs. Although careful navigation of the difficulties associated with data integration, validation, and regulatory routes is necessary, the potential benefits to patient outcomes and shortened development times make the effort valuable. Realizing the full potential of AI in medication repurposing will require a collaborative effort between clinical experience and computational approaches as technology progresses, ultimately improving the state of modern medicine. Accepting these advances will open the door for more individualized and successful treatment of complex diseases in addition to advancing therapeutic choices.

REFERENCES:

1. Lalu Muhammad Irham, Zalik Nuryana, Dyah Aryani Perwitasari, Yudha Rizky Nuari, Made Ary Sarasmita, Wirawan Adikusuma, Haafizah Dania, Rita Maliza, Rocky Cheung. Worldwide Publication Trends of Drug Repurposing and Drug Repositioning in the Science of Medicine (2003-2022). Research Journal of Pharmacy and Technology. 2023; 16(3): 1333-1. doi: 10.52711/0974-360X.2023.00219.

2. Sirota, M., et al., Discovery and preclinical validation of drug indications using compendia of public gene expression data. Science Translational Medicine. 2011. 3(96): p. 96ra77. https://doi.org/10.1126/scitranslmed.3001318.

3. Ayyar, P. and U. Subramanian, Repurposing – second life for drugs. Pharmacia. 2022. 69(1): p. 51-59. https://doi.org/10.3897/pharmacia.69.e72548.

4. Langedijk, J., et al., Drug repositioning and repurposing: terminology and definitions in literature. Drug Discovery Today, 2015. 20(8): p.1027 34 https://doi.org/10.1016/j.drudis.2015.05.001.

5. Sahragardjoonegani, B., et al., Repurposing existing drugs for new uses: a cohort study of the frequency of FDA-granted new indication exclusivities since 1997. Journal of Pharmaceutical Policy and Practice, 2021. 14(1): p. 3. https://doi.org/10.1186/s40545-020-00282-8.

6. Ayyar, P. and U. Subramanian, Repurposing – second life for drugs. Pharmacia, 2022. 69(1): p. 51-59. https://doi.org/10.3897/pharmacia.69.e72548.

7. Hay, M., et al., Clinical development success rates for investigational drugs. Nature Biotechnology, 2014. 32(1): p. 40-51. https://doi.org/10.1038/nbt.2786.

8. Annie Persaud. Using the EFQM Excellence Model Within Health Care, a Practical Guide to Success. International J. Health Care Quality Assurance. 2002; 15 (4): 141.

9. Charles A. Cianfrani, John E. ISO 9001: 2008 Explained, Third Edition, ASQ store; north America. P. 12.

10. Muralidharan, N., et al., Computational studies of drug repurposing and synergism of lopinavir, oseltamivir and ritonavir binding with SARS-CoV-2 protease against COVID-19. Journal of Biomolecular Structure and Dynamics. 2021. 39(7): p. 2673-2678. https://doi.org/10.1080/07391102.2020.1752802.

11. Pushpakom, S., et al., Drug repurposing: progress, challenges and recommendations. Nature Reviews. Drug Discovery, 2019. 18(1): p. 41-58. https://doi.org/10.1038/nrd.2018.168

12. Centers for Disease Control and Prevention (CDC). Update: Outbreak of severe acute respiratory syndrome--worldwide, 2003. MMWR Morb Mortal Wkly Rep. 2003; 52(12): 241-6.

13. Buchholz U, Müller MA, Nitsche A, Sanewski A, Wevering N, Bauer-Balci T, et al. Contact investigation of a case of human novel coronavirus infection treated in a German hospital, October-November 2012. Euro Surveill. 2013; 18: 20406.

14. Krishnamurthy, Nithya and Grimshaw, Alyssa and Axson, Sydney and Choe, Sung and Miller, Jennifer. (2021). Why promising drugs are shelved and barriers and facilitators to re-purposing them: A systematic literature review. 10.1101/2021.09.28.21264254.

15. Deng J, Yang Z, Ojima I, Samaras D, Wang F. Artificial intelligence in drug discovery: applications and techniques. Brief Bioinform. 2022; Jan 17; 23(1): bbab430. doi: 10.1093/bib/bbab430. PMID: 34734228.

Received on 09.10.2024 Revised on 30.12.2024

Accepted on 05.02.2025 Published on 28.02.2025

Available online from March 03, 2025

Asian J. Pharm. Res. 2025; 15(1):72-76.

DOI: 10.52711/2231-5691.2025.00012

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Creative Commons License.

Tool	Prediction/ identification	Approach	Script
Anni 2.0	Interprets differentially expressed genes and Literature based knowledge discovery	Concept based approach	UMLS and JAVA
Balestra Web	Interaction partners of any drug	Similarity Ensemble approach	Python, Flask, NumPy and SciPy
DrugQuest	Associations between known drugs	Clustering and test quest algorithms	CGI, PERL and JAVA
PolySearch	Human genes and diseases relationships	Simple dictionary approach	HTML and PERL