期刊
EXTREME MECHANICS LETTERS
卷 36, 期 -, 页码 -出版社
ELSEVIER
DOI: 10.1016/j.eml.2020.100652
关键词
Protein; Nanomechanics; Artificial intelligence; Machine learning; Deep neural networks; Folding; Structure prediction; Computation
资金
- IBM-MIT AI lab, United States of America
- Office of Naval Research (ONR), United States of America [N000141612333, N000141812258]
- National Institutes of Health (NIH), United States of America [U01 EB014976]
- Army Research Office (ARO), United States of America [73793EG]
- National Science Foundation [CMMI-1752172]
- U.S. Department of Defense (DOD) [N000141812258, N000141612333] Funding Source: U.S. Department of Defense (DOD)
The development of rational techniques to discover new mechanically relevant proteins for use in variety of applications ranging from mechanics, agriculture to biotechnology remains an outstanding nanomechanical design problem. The key barrier is to design a sequence to fold into a predictable structure to achieve a certain material function. Focused on alpha-helical proteins (as found in skin, hair, and many other mechanically relevant protein materials), we report a Multi-scale Neighborhood-based Neural Network (MNNN) model to learn how a specific amino acid sequence folds into a protein structure. The algorithm predicts the protein structure without using a template or co-evolutional information at a maximum error of 2.1 A. We find that the prediction accuracy is higher than other models and the prediction consumes less than six orders of magnitude time than ab initio folding methods. We demonstrate that MNNN can predict the structure of an unknown protein that agrees with experiments, and our model hence shows a great advantage in the rational design of de novo proteins. (c) 2020 Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据