Wang Yifan, Song Haitao, Teng Yue, Huang Guan, Qian Jingzhe, Wang Hongyu, Dong Shiyan, Ha JongHoon, Ma Yifan, Chang Mengyu, Jeong Seong Dong, Deng Weiye, Schrank Benjamin R, Grippin Adam, Wu Annette, Edwards Jared L, Zhang Yixiang, Lin Yuanyuan, Poon Wilson, Wilhelm Stefan, Bi Ye, Teng Lesheng, Wang Zikai, Kim Betty Y S, Jiang Wen
Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas 77030, United States.
SigmaSafari Project, Shanghai Artificial Intelligence Research Institute Co., Ltd., Shanghai 201109, China.
ACS Nano. 2025 May 27;19(20):19394-19407. doi: 10.1021/acsnano.5c03454. Epub 2025 May 14.
The recent success of large language models (LLMs) in performing natural language processing tasks has increased interest in applying generative artificial intelligence (AI) to scientific research. However, a common problem of LLMs is their tendency to produce inaccurate and sometimes "hallucinated" outputs. Here, we established a generative AI tool, NanoSafari, to automatically extract knowledge from the biomedical nanoscience literature and address scientific queries. We developed the Grouped Iterative Validation based Information Extraction (GIVE) method to extract contextual information on nanoparticle characteristics from >20,000 published articles and established a database that was incorporated into the generative LLM to provide accurate nanomaterial design parameters. Blinded evaluation by biomedical nanoscientists showed that NanoSafari outperformed the baseline model in providing more reliable parameters for nanomaterial design tasks, as further validated by bench experiments. Together, these findings demonstrate the utility of AI-based methods for automated learning from "real-world" published work to provide accurate and reliable scientific references for biomaterial and bioengineering applications.
大语言模型（LLMs）近期在执行自然语言处理任务方面取得的成功，引发了人们对将生成式人工智能（AI）应用于科学研究的兴趣。然而，大语言模型的一个常见问题是它们倾向于产生不准确且有时是“幻觉”的输出。在此，我们建立了一个生成式人工智能工具NanoSafari，用于自动从生物医学纳米科学文献中提取知识并回答科学问题。我们开发了基于分组迭代验证的信息提取（GIVE）方法，从20000多篇已发表文章中提取关于纳米颗粒特性的上下文信息，并建立了一个数据库，该数据库被纳入生成式大语言模型以提供准确的纳米材料设计参数。生物医学纳米科学家的盲法评估表明，NanoSafari在为纳米材料设计任务提供更可靠参数方面优于基线模型，这一点也得到了 bench实验的进一步验证。总之，这些发现证明了基于人工智能的方法在从“真实世界”已发表的工作中进行自动学习，为生物材料和生物工程应用提供准确可靠的科学参考方面的实用性。
