Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.

The topology of protein folds can be specified by the inter-residue contact-maps and accurate contact-map prediction can help ab initio structure folding. We developed TripletRes to deduce protein contact-maps from discretized distance profiles by end-to-end training of deep residual neural-networks...

Full description

Saved in:
Bibliographic Details
Main Authors: Yang Li, Chengxin Zhang, Eric W Bell, Wei Zheng, Xiaogen Zhou, Dong-Jun Yu, Yang Zhang
Format: Article
Language:English
Published: Public Library of Science (PLoS) 2021-03-01
Series:PLoS Computational Biology
Online Access:https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1008865&type=printable
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850239798555443200
author Yang Li
Chengxin Zhang
Eric W Bell
Wei Zheng
Xiaogen Zhou
Dong-Jun Yu
Yang Zhang
author_facet Yang Li
Chengxin Zhang
Eric W Bell
Wei Zheng
Xiaogen Zhou
Dong-Jun Yu
Yang Zhang
author_sort Yang Li
collection DOAJ
description The topology of protein folds can be specified by the inter-residue contact-maps and accurate contact-map prediction can help ab initio structure folding. We developed TripletRes to deduce protein contact-maps from discretized distance profiles by end-to-end training of deep residual neural-networks. Compared to previous approaches, the major advantage of TripletRes is in its ability to learn and directly fuse a triplet of coevolutionary matrices extracted from the whole-genome and metagenome databases and therefore minimize the information loss during the course of contact model training. TripletRes was tested on a large set of 245 non-homologous proteins from CASP 11&12 and CAMEO experiments and outperformed other top methods from CASP12 by at least 58.4% for the CASP 11&12 targets and 44.4% for the CAMEO targets in the top-L long-range contact precision. On the 31 FM targets from the latest CASP13 challenge, TripletRes achieved the highest precision (71.6%) for the top-L/5 long-range contact predictions. It was also shown that a simple re-training of the TripletRes model with more proteins can lead to further improvement with precisions comparable to state-of-the-art methods developed after CASP13. These results demonstrate a novel efficient approach to extend the power of deep convolutional networks for high-accuracy medium- and long-range protein contact-map predictions starting from primary sequences, which are critical for constructing 3D structure of proteins that lack homologous templates in the PDB library.
format Article
id doaj-art-ebd691bba94f4b35990eb6e0cb86a802
institution OA Journals
issn 1553-734X
1553-7358
language English
publishDate 2021-03-01
publisher Public Library of Science (PLoS)
record_format Article
series PLoS Computational Biology
spelling doaj-art-ebd691bba94f4b35990eb6e0cb86a8022025-08-20T02:01:03ZengPublic Library of Science (PLoS)PLoS Computational Biology1553-734X1553-73582021-03-01173e100886510.1371/journal.pcbi.1008865Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.Yang LiChengxin ZhangEric W BellWei ZhengXiaogen ZhouDong-Jun YuYang ZhangThe topology of protein folds can be specified by the inter-residue contact-maps and accurate contact-map prediction can help ab initio structure folding. We developed TripletRes to deduce protein contact-maps from discretized distance profiles by end-to-end training of deep residual neural-networks. Compared to previous approaches, the major advantage of TripletRes is in its ability to learn and directly fuse a triplet of coevolutionary matrices extracted from the whole-genome and metagenome databases and therefore minimize the information loss during the course of contact model training. TripletRes was tested on a large set of 245 non-homologous proteins from CASP 11&12 and CAMEO experiments and outperformed other top methods from CASP12 by at least 58.4% for the CASP 11&12 targets and 44.4% for the CAMEO targets in the top-L long-range contact precision. On the 31 FM targets from the latest CASP13 challenge, TripletRes achieved the highest precision (71.6%) for the top-L/5 long-range contact predictions. It was also shown that a simple re-training of the TripletRes model with more proteins can lead to further improvement with precisions comparable to state-of-the-art methods developed after CASP13. These results demonstrate a novel efficient approach to extend the power of deep convolutional networks for high-accuracy medium- and long-range protein contact-map predictions starting from primary sequences, which are critical for constructing 3D structure of proteins that lack homologous templates in the PDB library.https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1008865&type=printable
spellingShingle Yang Li
Chengxin Zhang
Eric W Bell
Wei Zheng
Xiaogen Zhou
Dong-Jun Yu
Yang Zhang
Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.
PLoS Computational Biology
title Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.
title_full Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.
title_fullStr Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.
title_full_unstemmed Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.
title_short Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks.
title_sort deducing high accuracy protein contact maps from a triplet of coevolutionary matrices through deep residual convolutional networks
url https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1008865&type=printable
work_keys_str_mv AT yangli deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks
AT chengxinzhang deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks
AT ericwbell deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks
AT weizheng deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks
AT xiaogenzhou deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks
AT dongjunyu deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks
AT yangzhang deducinghighaccuracyproteincontactmapsfromatripletofcoevolutionarymatricesthroughdeepresidualconvolutionalnetworks