ساخت اکانت کاربری رایگان

مقالات پذیرفته شده کنگره

  • صفحه نخست
  • مقالات پذیرفته شده اولین کنگره بین المللی فارماکوژنتیک
بازنشر

  • Artificial Intelligence and Large Language Models (LLMs) in Genomic Medicine, Bioinformatics, and Digital Health: Real-Time Monitoring of Patient Drug Response Using Big Data and Personalized Treatment

  • Kourosh Khayam Abed,1,* Sahar Masoomi,2
    1. Undergraduate Student in Biotechnology, Naqsh-e Jahan University, Isfahan, Iran
    2. Researcher in Medical Genetics, Pharmacogenomics, Pharmaceutical Biotechnology, Neuroscience, Nanomedicine, and Applied Artificial Intelligence in Personalized Medicine


  • Introduction: With the remarkable advancements in artificial intelligence and the emergence of Large Language Models (LLMs), the analysis of genomic data and the development of personalized medicine have entered a transformative stage. These models can process vast amounts of genomic, proteomic, and scientific text data, extract complex biological patterns, and leverage machine learning and deep learning algorithms to improve predictions of patient drug response. Simultaneously, the integration of digital health data and big data—including information from wearable devices, bioelectronic sensors, and electronic medical records—enables real-time monitoring of patient status and optimization of personalized pharmacotherapy. The application of LLMs in this domain, beyond analyzing scientific and genomic texts, offers the ability to integrate multi-source data and provide insights based on real-world evidence, supporting precise therapeutic interventions and rapid, effective clinical decision-making. This systematic review, focusing on the applications of artificial intelligence and large language models in genomic medicine and digital health, aims to provide a comprehensive overview of recent advancements, challenges, and opportunities in using big data for predicting drug response and developing personalized treatments. Such a perspective offers a novel outlook for precision and personalized medicine and highlights the potential of LLMs in transforming digital healthcare.
  • Methods: This systematic review examines studies published between 2020 and 2025 on the use of artificial intelligence, especially Large Language Models (LLMs), in genomics and digital health for personalized medicine. Major databases including PubMed, Scopus, Web of Science, SpringerLink, and arXiv were searched using keywords such as “AI,” “large language models,” “genomics,” “digital health,” and “personalized medicine.” Data were extracted on AI models, analyzed genomic or clinical datasets, predictive performance for drug response, integration with digital health or wearable devices, and reported limitations. The studies were synthesized to summarize trends, methodologies, and applications in AI-driven genomics and personalized medicine.
  • Results: Systematic analysis of studies published between 2020 and 2025 demonstrates that the integration of Large Language Models (LLMs) with deep multi-modal learning frameworks and multi-omics data analytics has generated unprecedented capacity for the identification of key genetic variants, predictive pharmacogenomic biomarkers, and complex gene regulatory networks. These models, leveraging transformer architectures and distributional data representations, can extract epigenetic and proteomic patterns associated with pharmacogenomics and enhance patient-specific drug response prediction. Integration of LLMs with digital health and big data, including wearable systems, bioelectronic sensors, and electronic health records, enables real-time monitoring of physiological and molecular biomarkers and assessment of dynamic cellular signaling networks. Multi-source integration and cross-modal deep learning analyses significantly enhance sensitivity and precision in pharmacogenomic predictions, minimizing potential errors in clinical decision-making. Moreover, evidence indicates that LLMs can extract insightful knowledge from real-world data, enabling the development of evidence-based decision support models and facilitating the design of personalized therapeutic interventions. Key challenges, however, remain in areas of model interpretability, training data bias, data security, and algorithm standardization.
  • Conclusion: This review demonstrates that the convergence of advanced artificial intelligence and large language models with multi-omics analytics and digital health data provides transformative capabilities for the design of personalized medicine and patient-specific drug response prediction. These models, by identifying functionally relevant genetic variants, predictive molecular biomarkers, and complex regulatory networks and signaling pathways, enable precise pharmacogenomic optimization and data-driven clinical decision-making. Integration with big data and advanced wearable systems allows dynamic, real-time monitoring of drug response, signaling network analysis, and extraction of predictive insights from real-world data, providing a robust framework for precision, personalized, and evidence-based medicine. Future directions include enhancing model interpretability, mitigating data biases, improving privacy and security standards, and implementing intelligent multi-source, multi-level algorithms. Ultimately, the combination of genomic analytics, deep learning, large language models, and digital health data paves the way for a fundamental transformation in personalized medicine, drug response prediction, and optimization of precision pharmacogenomic therapies, maximizing treatment efficacy and the overall impact of digital healthcare systems.
  • Keywords: Large Language Models (LLMs); Pharmacogenomics; Multi-omics; Digital Health; Personalized Medicine

به خانواده بزرگ فارماکوژنتیک و فارماکوژنومیکس بپیوندید!

ایجاد حساب کاربری رایگان