近日，来自山东大学、百图生科（BioMap）、北京理工大学、湖北医药学院、宁夏医科大学和阿卜杜拉国王科技大学（KAUST）的研究团队，在蛋白质研究领域取得重大突破。他们提出了一种全新的神经网络模型RMSF-net，能够在几秒钟内准确推断蛋白质的动力学信息。
蛋白质的动力学对于理解其机制至关重要，但预测蛋白质动学信息却极具挑战性。此次，研究团队通过深度学习和大规模蛋白质动力学数据集的整合，成功开发出RMSF-net模型。该模型能够整合复杂的原子结构和低温电子显微镜（cryo-EM）数据，准确识别两者之间的交互式双向约束和监督，从而提高动力学预测效率。
RMSF-net模型不仅预测准确，而且使用便捷。作为一个可免费使用的工具，它在蛋白质动力学研究中将发挥重要作用。研究团队以《Accurate Prediction of Protein Structural Flexibility by Deep Learning Integrating Intricate Atomic Structures and Cryo-EM Density Information》为题，于7月2日在《Nature Communication》上发表了相关研究成果。
该突破性的研究将为蛋白质研究和药物设计领域带来革命性的进展，有望为新型药物的研发提供更为精准的理论依据。
英语如下：
News Title: AI Model RMSF-net Accurately Predicts Protein Dynamics Information, Setting a New Industry Record
Keywords: 1. RMSF-net model
News Content:
Title: New Neural Network Model RMSF-net Makes a Quick and Accurate Prediction of Protein Dynamics Information
Recently, a research team from Shandong University, BioMap, Beijing Institute of Technology, Hubei University of Medicine, Ningxia Medical University, and King Abdullah University of Science and Technology (KAUST) has made a major breakthrough in protein research. They proposed a new neural network model, RMSF-net, which can accurately infer the dynamics information of proteins within seconds.
Protein dynamics are crucial to understanding its mechanisms, but predicting protein dynamics information is extremely challenging. This time, the research team successfully developed the RMSF-net model through the integration of deep learning and large-scale protein dynamics datasets. The model can integrate complex atomic structures and cryo-electron microscopy (cryo-EM) data, accurately identifying interactive bidirectional constraints and supervision between them, thereby improving the efficiency of dynamics prediction.
The RMSF-net model is not only accurate in prediction but also convenient to use. As a free tool, it will play an important role in protein dynamics research. The research team published relevant research results on July 2nd with the title “Accurate Prediction of Protein Structural Flexibility by Deep Learning Integrating Intricate Atomic Structures and Cryo-EM Density Information” in Nature Communication.
This groundbreaking research will bring revolutionary progress to the fields of protein research and drug design, and is expected to provide more accurate theoretical evidence for the development of new drugs.
【来源】https://www.jiqizhixin.com/articles/2024-07-05-5
Views: 3