ResPRE: high-accuracy protein contact prediction by coupling precision matrix with deep residual neural networks

By Yang Li, Junda Hu, Chengxin Zhang, Dong-Jun Yu, Yang Zhang. 2019

146 citations recorded by the source corpus.

MOTIVATION Contact-map of a protein sequence dictates the global topology of structural fold. Accurate prediction of the contact-map is thus essential to protein 3D structure prediction, which is particularly useful for the protein sequences that do not have close homology templates in the PDB. RESULTS We developed a new method, ResPRE, to predict residue-level protein contacts using inverse covariance matrix (or precision matrix) of multiple sequence alignments (MSAs) through deep residual convolutional neural network training. The approach was tested on a set of 158 non-homologous proteins collected from the CASP experiments and achieved an average accuracy of 50.6% in the top-L long-range contact prediction with L being the sequence length, which is 11.7% higher than the best of other state-of-the-art approaches ranging from coevolution coupling analysis to deep neural network training. Detailed data analyses show that the major advantage of ResPRE lies at the utilization of precision matrix that helps rule out transitional noises of contact-maps compared to the previously used covariance matrix. Meanwhile, the residual network with parallel shortcut layer connections increases the learning ability of deep neural network training. It was also found that appropriate collection of MSAs can further improve the accuracy of final contact-map predictions. The standalone package and online server of ResPRE are made freely available, which should bring important impact on protein structure and function modeling studies in particular for the distant- and non-homology protein targets. AVAILABILITY https://zhanglab.ccmb.med.umich.edu/ResPRE and https://github.com/leeyang/ResPRE. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Subfields: Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Assessing the accuracy of contact predictions in CASP13 (2019, background): “Despite the assumption that the input coupling matrix is random, 62 43 While more sequences have a positive impact on the accuracy of contact prediction in general, the deep residual neural networks have the ability to learn the underlying contact patterns from limited coevolutionary information;…”
Toward the solution of the protein structure prediction problem (2021, methodology): “C-I-TASSER C-I-TASSER and C-QUARK were extensions of the classic I-TASSER and QUARK frameworks, which included contact maps from TripletRes, ResPRE, and numerous deep learning-based predictors into their simulations.”
Deep-learning contact-map guided protein structure prediction in CASP13 (2019, background): “Second, the capability of NeBcon for accurate contact-map prediction was extended by integrating nine state-of-the-art contact prediction programs, including several deep-learning-based methods, such as ResPRE, 17 which is a newly developed contact-map predictor that couples coevolution-derived prec”
Deep learning techniques have significantly impacted protein structure prediction and protein design. (2021, methodology): “The approach introduced by RaptorX-Contact was adapted by methods such as ResPRE [28] and TripletRes [31 ], which use a similar deep learning architecture but with a unique set of features that include multiple co-evolutionary coupling matrices directly deduced from deep MSAs without…”
Ensembling multiple raw coevolutionary features with deep residual neural networks for contact-map prediction in CASP13 (2019, methodology): “Due to its significant importance to protein structure prediction, the problem of contact-map prediction has drawn an increasing amount of attention, and as a result, several new approaches have been proposed within the last decade.(10-21) Initial studies on protein contact prediction focused on the”
I-TASSER-MTD: a deep-learning-based platform for multi-domain protein structure and function prediction (2022, methodology): “Then the secondary structure is predicted by PSSpred 55 , and the contact map is generated by the deep-learning-based predictor ResPRE 10 , which has been included in DeepPotential, using features directly extracted from the MSA.”
Folding non-homologous proteins by coupling deep-learning contact maps with I-TASSER assembly simulations (2021, methodology): “Most recently, with the development of advanced algorithms in co-evolutionary decoupling (Jones et al., 2012; Marks et al., 2011; Weigt et al., 2009) and deep machine learning (Li et al., 2019a; Wang et al., 2017), contact prediction accuracy has dramatically increased.”
De novo Protein Structure Prediction by Coupling Contact With Distance Profile (2020, methodology): “In ResPRE [25], a coupling precision matrix with deep residual neural networks is used to predict protein contacts.”
Deducing high-accuracy protein contact-maps from a triplet of coevolutionary matrices through deep residual convolutional networks (2020, methodology): “In general, the COV-based model has the lowest precision among the three individual feature models, probably due to the translational noise in the covariance matrix [27].”
Integrating end-to-end learning with deep geometrical potentials for ab initio RNA structure prediction (2022, unknown): “The length range of the test RNAs [14, 392 NTs] was slightly larger than but generally consistent with the crop-size range of the RNAs that were used for training the DRfold models (i.e., <200 NTs) due to the GPU memory limitations.”
FARFAR2: Improved de novo Rosetta prediction of complex global RNA folds (2020, unknown): “modeling pools for groups developing complementary procedures, including recent scoring methods taking advantage of artificial neural networks and evolutionary coupling information (Li et al., 2019; Wang et al., 2019; Weinreb et al., 2016) and alternative highresolution refinement procedures (Tan et”