{
  "id": 20845430,
  "title": "AI virtual cell technology could reveal how genetic changes and drugs alter cells",
  "url": "https://urgent.news/2026/11/20/ai-virtual-cell-technology-could-reveal-how-genetic-changes-and-drugs",
  "topic": "ai",
  "section": "AI",
  "published": "2026-11-20T03:00:06.000Z",
  "source": {
    "name": "Phys.org",
    "slug": "phys-org",
    "url": "https://phys.org/news/2026-11-ai-virtual-cell-technology-reveal.html"
  },
  "original_language": "en",
  "account": "A new artificial intelligence (AI) virtual cell technology developed by researchers at Texas A&M University aims to shed light on the underlying mechanisms of how genetic alterations and drug treatments influence cellular behavior. The innovative system, named PertReason, has the potential to generate a more comprehensive understanding of disease pathways and identify novel drug targets for therapeutic intervention.\n\nLed by Dr. Yang Shen, a professor in the Electrical and Computer Engineering department, the project involves postdoctoral researcher Dr. Dongkwan Kim, Ph.D. student Yiming Gao, and Yining Yang, another Ph.D. student. The team's current model has been accepted for presentation at the upcoming Conference on Neural Information Processing Systems (NeurIPS 2026) and is already available on the arXiv preprint server. A previous version was showcased at the 3rd Workshop on Multi-modal Foundation Models and Large Language Models for Life Sciences during the International Conference on Machine Learning (ICML 2026) in Seoul, South Korea.\n\nUnlike conventional approaches that focus on individual biological components, PertReason can trace the effects of genetic and drug-induced alterations within cellular networks. By expanding the scale of study to the cellular level, researchers can now better comprehend why specific changes lead to diseases and explore more effective treatment strategies.\n\nTo ensure the accuracy and reliability of AI models, Shen's team developed a benchmark called PertReasonQA, which evaluates not only the correctness of predictions but also the rationale behind them. This tool helps identify instances where the model provides the right answer but fails to utilize the correct data in its reasoning process. Additionally, the researchers have trained an AI model to reason through biological pathways, enhancing the trustworthiness of its predictions.\n\nEarly in the research, the team collaborated with clinical scientists working with real breast cancer patients, gaining insights into the challenges faced by patients who initially respond to treatment before subsequently developing resistance. The team recognized the need to bridge the gap between AI model performance and real-world impact. By pinpointing the most promising biological pathways for further investigation, PertReason could streamline the discovery process, minimizing trial-and-error and accelerating medical breakthroughs.\n\nWith Texas A&M's new supercomputer VISION now available, Shen and his team plan to develop digital twins—highly detailed virtual replicas of the human body—using the enhanced capabilities of the AI virtual cell model. This ambitious endeavor aims to facilitate personalized medicine and support clinicians in making informed decisions about patient care. Ultimately, the goal is to harness AI to uncover new ways of understanding and treating diseases on a personalized level, leading to improved patient outcomes.",
  "summary": "Today's artificial intelligence models can predict whether someone with a genetic change will develop a disease, but cannot answer two critical questions: Why did the disease occur, and how do we design a treatment?",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}