Deep template-based protein structure prediction
文献类型：期刊论文
| 作者 | Wu, Fandi1,2,3; Xu, Jinbo3 |
| 刊名 | PLOS COMPUTATIONAL BIOLOGY
 |
| 出版日期 | 2021-05-01 |
| 卷号 | 17期号:5页码:18 |
| ISSN号 | 1553-734X |
| DOI | 10.1371/journal.pcbi.1008954 |
| 英文摘要 | Author summary TBM (template-based modeling) is a popular method for protein structure prediction. However, existing methods cannot generate good models when the protein under prediction does not have very similar templates in Protein Data Bank (PDB). Recently significant progress has been made on template-free protein structure prediction by deep learning, but very few deep learning methods were developed for TBM. To further improve TBM for protein structure prediction, we present a new deep learning method that greatly outperforms existing ones in identifying the best templates, generating sequence-template alignment and constructing 3D models from alignments. Blindly tested in CASP14, our server obtained the best average model quality score on the 58 TBM targets among all the CASP14-participating servers, which confirms that our method is effective for TBM. Motivation Protein structure prediction has been greatly improved by deep learning, but most efforts are devoted to template-free modeling. But very few deep learning methods are developed for TBM (template-based modeling), a popular technique for protein structure prediction. TBM has been studied extensively in the past, but its accuracy is not satisfactory when highly similar templates are not available. Results This paper presents a new method NDThreader (New Deep-learning Threader) to address the challenges of TBM. NDThreader first employs DRNF (deep convolutional residual neural fields), which is an integration of deep ResNet (convolutional residue neural networks) and CRF (conditional random fields), to align a query protein to templates without using any distance information. Then NDThreader uses ADMM (alternating direction method of multipliers) and DRNF to further improve sequence-template alignments by making use of predicted distance potential. Finally, NDThreader builds 3D models from a sequence-template alignment by feeding it and sequence coevolution information into a deep ResNet to predict inter-atom distance distribution, which is then fed into PyRosetta for 3D model construction. Our experimental results show that NDThreader greatly outperforms existing methods such as CNFpred, HHpred, DeepThreader and CEthreader. NDThreader was blindly tested in CASP14 as a part of RaptorX server, which obtained the best average GDT score among all CASP14 servers on the 58 TBM targets. |
| 资助项目 | National Institutes of Health[R01GM089753] ; National Science Foundation[DBI1564955] ; CSC Scholarship ; National Key Research and Development Program of China[2020AAA0103802] ; NSF of China[U20A20227] |
| WOS研究方向 | Biochemistry & Molecular Biology ; Mathematical & Computational Biology |
| 语种 | 英语 |
| 出版者 | PUBLIC LIBRARY SCIENCE |
| WOS记录号 | WOS:000646386700006 |
| 源URL | [http://119.78.100.204/handle/2XEOYT63/17815]  |
| 专题 | 中国科学院计算技术研究所期刊论文_英文 |
| 通讯作者 | Xu, Jinbo |
| 作者单位 | 1.Univ Chinese Acad Sci, Beijing, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China 3.Toyota Technol Inst, Chicago, IL 60637 USA |
| 推荐引用方式 GB/T 7714 | Wu, Fandi,Xu, Jinbo. Deep template-based protein structure prediction[J]. PLOS COMPUTATIONAL BIOLOGY,2021,17(5):18. |
| APA | Wu, Fandi,&Xu, Jinbo.(2021).Deep template-based protein structure prediction.PLOS COMPUTATIONAL BIOLOGY,17(5),18. |
| MLA | Wu, Fandi,et al."Deep template-based protein structure prediction".PLOS COMPUTATIONAL BIOLOGY 17.5(2021):18. |
入库方式： OAI收割
来源：计算技术研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。