新华社电 记者从中国药科大学获悉，该校多靶标天然药物全国重点实验室肖易倍教授团队近日揭示了 细菌通过代谢抵抗噬菌体感染的免疫新机制，为今后开发相关药物提供了思路。 肖易倍介绍，噬菌体是一类专门感染细菌的病毒。近年来，国内外研究发现，作为一种单细胞生物，细 菌竟能够抑制噬菌体的感染和传播。"以往科学界认为，只有人类这样的高级动物才拥有免疫系统，细 菌抗病毒的机制和免疫系统很像，因此被称为'细菌免疫'。" 团队成员、中国药科大学药学院副教授陈美容告诉记者，此次研究成果基于CRISPR-Cas系统，基因编 辑技术就来自CRISPR-Cas系统的一个分型，该技术就像剪刀，能够将遗传物质从特定位置切断。 转自：贵州日报 "此前有研究显示，细菌被噬菌体侵染后，会激活体内的Ⅲ型CRISPR-Cas系统，通过切割噬菌体的遗传 物质，干扰其复制。"陈美容介绍，团队历经两年多研究，发现了另一种基于ATP代谢的免疫新机制。 "这是一招'釜底抽薪'，也就是把细菌体内的能量因子ATP消耗殆尽。"中国药科大学生命科学与技术学 院副教授陆美玲说，"生命活动需要能量，这种新机制把ATP代谢为具有毒性的ITP。噬菌体缺少足够能 量进行自我复制 ...

Core Viewpoint - Fudan University has made significant breakthroughs in the medical field, including the discovery of a new treatment target for Parkinson's disease and the upcoming launch of early screening and diagnostic testing for Alzheimer's disease, supported by AI computing power from the CFFF platform in collaboration with Alibaba Cloud [1][3]. Group 1: Breakthroughs in Alzheimer's and Parkinson's Disease - The research team led by Professor Yu Kintai has achieved a 15-year early prediction of Alzheimer's disease risk with over 98.7% accuracy, published in the journal Nature [3][5]. - The team has identified a new treatment target for Parkinson's disease and utilized AI to screen potential drugs, with findings published in top international journals such as Cell and Nature [3][5]. Group 2: CFFF Platform and AI Integration - The CFFF platform, launched in 2023, integrates advanced computing clusters and AI technologies, enabling researchers to process large datasets more efficiently than traditional methods [3][4]. - The platform supports over 1,000 parallel intelligent computations and facilitates the training of large models with billions of parameters, significantly enhancing research capabilities [3][4]. Group 3: Efficiency Improvements in Research - The use of AI and innovative data-driven methods has allowed the research team to analyze over 6,361 cerebrospinal fluid proteins, identifying five key proteins that improve diagnostic accuracy to 98.7% [4][5]. - AI technology has accelerated the identification of potential therapeutic targets in Parkinson's disease, completing in five years what would traditionally take decades [5].

Core Viewpoint - UK BioBank emphasizes the importance of its database for global health and disease research, particularly in light of recent restrictions imposed by the NIH on data access for researchers in China and other countries [1][4][5]. Group 1: Impact of NIH Restrictions - The NIH has prohibited access to its controlled databases, including the SEER database, for researchers from specific countries, including China, effective April 4, 2025 [4][8]. - The SEER database is crucial for cancer research, covering data from 48% of the U.S. population, and has been relied upon by approximately 75% of cancer epidemiology papers published by Chinese scholars [9][10]. - The restrictions have raised concerns about a potential "cold war" in scientific research, with fears that other databases may follow suit in limiting access for Chinese researchers [5][10]. Group 2: Need for Domestic Data Sharing - Chinese research institutions must enhance their capabilities and promote scientific data sharing to avoid being significantly hindered by international restrictions [6][10]. - The current state of biomedical research in China shows a significant reliance on foreign databases, with 99% of pharmaceuticals and 100% of databases not being domestically sourced [13][14]. - Since 2004, China has made efforts to build national scientific data sharing platforms, but challenges remain in the implementation and effectiveness of these initiatives [15][17]. Group 3: Challenges in Data Sharing - There is a systemic lack of focus on scientific data development in China, with many efforts being limited to small teams rather than a cohesive national strategy [17][18]. - The reliance on foreign databases in educational institutions hampers the development of domestic data products, which, despite having competitive potential, struggle due to low user engagement [18][19]. - The phenomenon of "false sharing" is prevalent, where databases are claimed to be open but are not genuinely accessible, leading to a cycle of underutilization and slow development [21][22].