晨间信号Morning Signal
《科学》 第392卷 第6793期 · 2026年4月2日 · 中文解读

防御预测器:发现原核免疫系统的机器学习模型

DefensePredictor: A machine learning model to discover prokaryotic immune systems
约 76 分钟Research在小程序里点播,20 到 60 分钟做好
这篇讲什么
本文介绍了一种名为DefensePredictor的机器学习模型,该模型利用蛋白质语言模型嵌入来分类蛋白质是否具有防御功能,并在大肠杆菌和多种原核基因组中预测和验证了新的抗噬菌体防御系统。
原文开头
Full article and list of author affiliations: https://doi.org/10.1126/ science.adv7924 Peter C. DeWeirdt, Emily M. Mahoney, Michael T. Laub* INTRODUCTION: Bacteriophages (viruses that infect bacteria) are the most abundant biological entity on the planet. To withstand phage infection, bacteria have developed sophisticated innate immune systems that are analogous, and sometimes directly homologous, to eukaryotic innate immune systems. Because of their molecular specificity, some of these prokaryotic immune systems, such as CRISPR-Cas and restriction enzymes, have been repurposed as powerful biotechnologies. Dozens of new prokaryotic immune systems have been discovered in recent years, but the full repertoire remains unknown. RATIONALE: Antiphage immunity genes often cluster in prokaryotic genomes in so-called “defense islands.” …
摘自《科学》(Science)第392卷 第6793期 · 2026年4月2日。仅引用开头一小段供了解文章,版权归原刊所有,全文请阅读原刊。
晨间信号小程序码
微信扫码,在小程序里听完整版
不用登录先听一篇 · 或在微信搜索小程序 晨间信号
同期其他文章