Improved protein structure prediction using potentials from deep learning

doi:10.1038/s41586-019-1923-7

GSTDTAP > 地球科学

DOI	10.1038/s41586-019-1923-7
	Improved protein structure prediction using potentials from deep learning
	Ma, Runze 1,2; Cao, Duanyun 1; Zhu, Chongqin 3; Tian, Ye 1; Peng, Jinbo 1; Guo, Jing 1; Chen, Ji 5; Li, Xin-Zheng 5,6; Francisco, Joseph S.3; Zeng, Xiao Cheng 4,7,8,9,10; Xu, Li-Mei 1,6; Wang, En-Ge 1,11,12; Jiang, Ying 1,6,13
	2020-01-02
发表期刊	NATURE
ISSN	0028-0836
EISSN	1476-4687
出版年	2020
卷号	577 期号:7792 页码:706-+
文章类型	Article
语种	英语
国家	England
英文关键词	Protein structure prediction can be used to determine the three-dimensional shape of a protein from its amino acid sequence(1). This problem is of fundamental importance as the structure of a protein largely determines its function(2) however, protein structures can be difficult to determine experimentally. Considerable progress has recently been made by leveraging genetic information. It is possible to infer which amino acid residues are in contact by analysing covariation in homologous sequences, which aids in the prediction of protein structures(3). Here we show that we can train a neural network to make accurate predictions of the distances between pairs of residues, which convey more information about the structure than contact predictions. Using this information, we construct a potential of mean force(4) that can accurately describe the shape of a protein. We find that the resulting potential can be optimized by a simple gradient descent algorithm to generate structures without complex sampling procedures. The resulting system, named AlphaFold, achieves high accuracy, even for sequences with fewer homologous sequences. In the recent Critical Assessment of Protein Structure Prediction(5) (CASP13)-a blind assessment of the state of the field-AlphaFold created high-accuracy structures (with template modelling (TM) scores(6) of 0.7 or higher) for 24 out of 43 free modelling domains, whereas the next best method, which used sampling and contact information, achieved such accuracy for only 14 out of 43 domains. AlphaFold represents a considerable advance in protein-structure prediction. We expect this increased accuracy to enable insights into the function and malfunction of proteins, especially in cases for which no structures for homologous proteins have been experimentally determined(7).
领域	地球科学 ; 气候变化 ; 资源环境
收录类别	SCI-E
WOS记录号	WOS:000510520700067
WOS关键词	SECONDARY STRUCTURE ; NEURAL-NETWORKS ; CONTACTS ; COEVOLUTION ; SEQUENCES
WOS类目	Multidisciplinary Sciences
WOS研究方向	Science & Technology - Other Topics
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.173/C666/handle/2XK7JSWQ/281376
专题	地球科学资源环境科学气候变化
作者单位	1.Peking Univ, Sch Phys, Int Ctr Quantum Mat, Beijing, Peoples R China; 2.Huairou Natl Comprehens Sci Ctr, Phys Sci Lab, Beijing, Peoples R China; 3.Univ Penn, Dept Earth & Environm Sci, Philadelphia, PA 19104 USA; 4.Univ Nebraska, Dept Chem, Lincoln, NE 68588 USA; 5.Peking Univ, Sch Phys, Beijing, Peoples R China; 6.Collaborat Innovat Ctr Quantum, Beijing, Peoples R China; 7.Univ Nebraska, Dept Chem & Biomol Engn, Lincoln, NE 68588 USA; 8.Univ Nebraska, Dept Mech & Mat Engn, Lincoln, NE 68588 USA; 9.Univ Nebraska, Dept Phys, Lincoln, NE 68588 USA; 10.Univ Nebraska, Nebraska Ctr Mat & Nanosci, Lincoln, NE 68588 USA; 11.Chinese Acad Sci, Inst Phys, Songshan Lake Mat Lab, Ceram Div, Guangzhou, Guangdong, Peoples R China; 12.Liaoning Univ, Sch Phys, Shenyang, Liaoning, Peoples R China; 13.Univ Chinese Acad Sci, CAS Ctr Excellence Topol Quantum Computat, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Ma, Runze,Cao, Duanyun,Zhu, Chongqin,et al. Improved protein structure prediction using potentials from deep learning[J]. NATURE,2020,577(7792):706-+.
APA	Ma, Runze.,Cao, Duanyun.,Zhu, Chongqin.,Tian, Ye.,Peng, Jinbo.,...&Jiang, Ying.(2020).Improved protein structure prediction using potentials from deep learning.NATURE,577(7792),706-+.
MLA	Ma, Runze,et al."Improved protein structure prediction using potentials from deep learning".NATURE 577.7792(2020):706-+.