INTERNATIONAL CONGRESS OF
PharmacoGenetics
AI-Driven Pharmacogenomics in Oncology: Overcoming Drug Resistance in Cancer Therapy
-
Saghar Modaresi,1,*
1. Independent Researcher, B.Sc. in Laboratory Sciences
- Introduction: Drug resistance is a critical challenge in cancer treatment, often leading to therapy failure, relapse, and poor patient outcomes. Conventional therapies, including chemotherapy, radiotherapy, and immunotherapy, are frequently limited by the emergence of drug-resistant tumor phenotypes. Pharmacogenomics (PGx) studies patient-specific genetic determinants that influence drug response, offering a path toward personalized medicine. Recently, artificial intelligence (AI) has been integrated with multi-omics data—including genomics, transcriptomics, proteomics, metabolomics, and epigenomics—to address the complexity of drug resistance. AI techniques such as machine learning (ML) and deep learning (DL) enable the identification of hidden patterns in large datasets, prediction of therapeutic response, and design of optimized treatment strategies. This review summarizes current advancements in AI-driven pharmacogenomics for understanding, predicting, and overcoming drug resistance in oncology.
- Methods: A comprehensive literature search was conducted in PubMed, Web of Science, Embase, and Cochrane databases up to 2025. Keywords included “artificial intelligence,” “machine learning,” “deep learning,” “pharmacogenomics,” “drug resistance,” “cancer therapy,” and “multi-omics.” Inclusion criteria encompassed English-language studies published in the last six years that addressed AI applications for predicting or analyzing cancer drug resistance. The Prediction Model Risk of Bias Assessment tool was used to evaluate the quality of studies. Extracted data included cancer types, AI algorithms, omics datasets, predictive accuracy, and clinical relevance.
- Results: AI-based pharmacogenomics has shown substantial potential in predicting drug resistance and guiding therapeutic decisions. Integration of multi-omics datasets allows for the elucidation of complex resistance mechanisms, including gene mutations, overexpression of drug efflux pumps, epigenetic modifications, and tumor microenvironment alterations such as hypoxia and immune evasion. Deep learning models, including DrugS, DrugnomeAI, and PandaOmics, have been employed to predict drug response, optimize combination therapies, and stratify patients based on expected outcomes. In vitro and in vivo validation, including patient-derived xenografts, demonstrated that AI-guided therapy can effectively overcome resistance, for example using inhibitors of CDK, mTOR, or apoptosis pathways. Large-scale datasets, such as The Cancer Genome Atlas (TCGA), have been leveraged to enhance population-level insights and support precision oncology strategies. Nonetheless, challenges persist, including data heterogeneity, model interpretability, ethical considerations, and integration into clinical workflows.
- Conclusion: The integration of AI and pharmacogenomics presents a transformative approach for addressing drug resistance in cancer therapy. By combining multi-omics data with advanced computational models, AI enables the identification of resistance mechanisms, prediction of patient-specific drug responses, and optimization of treatment regimens. While current research demonstrates promising results, further work is needed to enhance model interpretability, ensure regulatory compliance, and validate findings in clinical settings. Continued advancement in AI-driven pharmacogenomics has the potential to significantly improve personalized cancer treatment, reduce therapy failure, and enhance patient survival and quality of life.
- Keywords: Artificial intelligence; pharmacogenomics; cancer; drug resistance; multi-omics; precision medicine
