AI-powered drug repurposing: A game changer for cancer research

The field of cancer treatment has long struggled with the immense costs and time-consuming nature of drug development. Traditional methods often take over a decade and billions of dollars to bring a single drug to market, with many compounds failing in late-stage trials due to efficacy or safety concerns. However, artificial intelligence (AI) is now revolutionizing this space by accelerating drug repurposing and designing new therapeutics with unprecedented speed and accuracy. The integration of AI in oncology drug discovery holds the promise of reducing development timelines, optimizing existing drugs, and unveiling novel treatment strategies.

A recent study titled Artificial Intelligence-Based Methods for Drug Repurposing and Development in Cancer, authored by Sara Herráiz-Gil, Elisa Nygren-Jiménez, Diana N. Acosta-Alonso, Carlos León, and Sara Guerrero-Aspizua, and published in Applied Sciences (2025, 15, 2798), presents an in-depth review of AI-driven techniques in drug discovery. The study highlights AI’s role in addressing critical challenges in oncology and explores the latest methodologies and applications in the field.

Role of AI in drug repurposing and new drug development

AI has emerged as a game-changer in the pharmaceutical industry, particularly in oncology, by offering powerful tools for drug repurposing and de novo drug design. Traditional drug repurposing involves identifying new uses for existing drugs, but AI significantly enhances this process by analyzing large-scale biological and chemical data. Machine learning and deep learning algorithms can predict drug-disease interactions, optimize drug efficacy, and minimize toxicity concerns.

The study discusses how knowledge graphs and neural networks are being employed to map complex relationships between drugs and diseases. Graph-based AI techniques allow researchers to identify potential drug candidates based on existing biological networks, while deep learning models can assess molecular interactions with remarkable precision. Generative AI models, such as reinforcement learning algorithms, are also gaining traction in de novo drug design, creating entirely new molecular structures optimized for cancer treatment. By leveraging multi-omics data, AI enables a more precise and personalized approach to therapy.

AI applications in cancer drug discovery

Several groundbreaking AI-driven projects have demonstrated the potential of this technology in oncology. The study outlines multiple case studies where AI was successfully applied to identify promising drug candidates. For instance, AI models have predicted potential therapies for chondrosarcoma, familial melanoma, and liver and lung cancers. By integrating diverse datasets, such as genomic profiles, protein interactions, and clinical trial results, these models provide insights into drug repositioning opportunities that might otherwise remain undiscovered.

Furthermore, AI has accelerated drug screening by predicting the binding affinity of molecules to cancer targets, reducing the need for costly and time-intensive laboratory testing. In hepatocellular carcinoma research, AI-driven de novo drug design has led to the development of a novel CDK20 inhibitor in record time, highlighting the efficiency of computational drug discovery. Additionally, AI-guided strategies have been instrumental in predicting patient responses to specific treatments, paving the way for more targeted and effective cancer therapies.

Experimental integration and challenges in AI-driven drug discovery

While AI presents remarkable opportunities, its integration into traditional experimental workflows comes with challenges. One major limitation is data quality - AI models rely on vast amounts of biomedical data, which can sometimes be inconsistent or biased. Ensuring data standardization and accuracy remains a critical challenge in training reliable AI models.

Another concern is the interpretability of AI predictions. Many deep learning models operate as “black boxes,” making it difficult for researchers to understand the rationale behind certain drug recommendations. To address this, explainable AI (XAI) techniques, such as SHAP and LIME, are being developed to enhance model transparency and regulatory acceptance. Ethical considerations, such as ensuring equitable access to AI-driven therapies and mitigating algorithmic biases, are also central to the responsible adoption of AI in drug discovery.

Despite these challenges, the study emphasizes that AI’s integration with experimental methods - such as AI-guided high-throughput screening, in vitro and in vivo testing, and AI-assisted synthesis - has the potential to overcome traditional bottlenecks in drug development. By continuously refining AI methodologies and fostering collaboration between academia, industry, and regulatory bodies, AI-driven drug discovery could redefine the future of oncology treatments.

Future prospects and conclusion

The convergence of AI, big data, and computational biology is ushering in a new era of precision oncology. With AI’s ability to analyze multi-omics datasets and predict drug interactions with high accuracy, the pharmaceutical landscape is witnessing a shift towards more efficient, cost-effective, and patient-specific cancer treatments. As AI continues to evolve, its role in drug discovery will expand further, with advancements in quantum computing and multimodal AI offering even greater potential.

Overall, the study underscores AI’s transformative impact on oncology drug discovery. While challenges remain, the ongoing advancements in AI-driven methodologies hold the promise of significantly improving cancer treatment outcomes. By bridging the gap between computational power and experimental validation, AI is not only accelerating drug discovery but also making personalized medicine a tangible reality for cancer patients worldwide.