CONFOLD: Residue-Residue Contact-guided ab initio Protein Folding

By Badri Adhikari, Debswapna Bhattacharya, Renzhi Cao, Jianlin Cheng. 2015

147 citations recorded by the source corpus.

Predicted protein residue–residue contacts can be used to build three‐dimensional models and consequently to predict protein folds from scratch. A considerable amount of effort is currently being spent to improve contact prediction accuracy, whereas few methods are available to construct protein tertiary structures from predicted contacts. Here, we present an ab initio protein folding method to build three‐dimensional models using predicted contacts and secondary structures. Our method first translates contacts and secondary structures into distance, dihedral angle, and hydrogen bond restraints according to a set of new conversion rules, and then provides these restraints as input for a distance geometry algorithm to build tertiary structure models. The initially reconstructed models are used to regenerate a set of physically realistic contact restraints and detect secondary structure patterns, which are then used to reconstruct final structural models. This unique two‐stage modeling approach of integrating contacts and secondary structures improves the quality and accuracy of structural models and in particular generates better β‐sheets than other algorithms. We validate our method on two standard benchmark datasets using true contacts and secondary structures. Our method improves TM‐score of reconstructed protein models by 45% and 42% over the existing method on the two datasets, respectively. On the dataset for benchmarking reconstructions methods with predicted contacts and secondary structures, the average TM‐score of best models reconstructed by our method is 0.59, 5.5% higher than the existing method. The CONFOLD web server is available at http://protein.rnet.missouri.edu/confold/. Proteins 2015; 83:1436–1449. © 2015 Wiley Periodicals, Inc.

Subfields: Materials Chemistry, 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 Materials Chemistry

In Molecular Biology

Citations in the corpus

Accurate De Novo Prediction of Protein Contact Map by Ultra-Deep Learning Model (2016, background): “the two residues in a contact falls into [6, 11], [12, 23], and ≥24, respectively.”
End-to-end differentiable learning of protein structure (2018, methodology): “Table 2 compares training and prediction speeds of RGNs to established methods that rely heavily on simulation with limited learning (first row), and co-evolution-based contact prediction methods that rely on learning (second row), combined with CONFOLD (Adhikari et al., 2015) to convert predicted c”
Enhancing Evolutionary Couplings with Deep Convolutional Neural Networks (2017, methodology): “We ran an off-the-shelf folding algorithm, Confold, without any further refinement and selected the top 5 of 500 models by energy for each target (Adhikari et al., 2015).”
Folding membrane proteins by deep transfer learning (2017, methodology): “Although so many methods have been developed, for MPs with fewer sequence homologs, the predicted contacts by existing methods are still of low quality (Adhikari et al., 2015).”
Large-scale structure prediction by improved contact predictions and model quality assessment (2017, methodology): “In contrast to PconsFold (Michel et al., 2014) PconsFold2 uses CONFOLD (Adhikari et al., 2015), i.e. the NMR protocol of CNS (Brunger, 2007) and not ROSETTA (Leaver-Fay et al., 2011) to generate 50 models for each contact map.”
CoinFold: a web server for protein contact prediction and contact-assisted protein folding (2016, methodology): “As shown in Figure 2A and C, our server can generate much better Top1 3D models on CASP and CAMEO hard targets than ConFold and EVfold.”
SE(3)-Equivariant Energy-based Models for End-to-End Protein Folding (2021, background): “There has been a remarkable progress towards more accurate protein structure prediction, largely owing to more precise inter-residue contact/distance and orientation predictions by deep networks [1, 2, 46, 48, 29, 30, 17].”
Predicting membrane protein contacts from non-membrane proteins by deep transfer learning (2017, methodology): “Although so many methods have been developed, for MPs with fewer sequence homologs, the predicted contacts by existing methods such as CCMpred, PSICOV, Evfold, MetaPSICOV and CoinDCA [18] are still of low quality and not very helpful to ab initio folding [19].”
Protein single-model quality assessment by feature-based probability density functions (2016, unknown): “polar residue i, and �� is the total mass of residue i. 7. The exposed surface score describes the percentage of area of the residues exposed, and is calculated as follows: �� 𝑎��= ∑�� ∑�� (7) �� is the total area of residue i parsed by DSSP, and �� is the exposed area of residue i. 8. ”