Trends in template/fragment-free protein structure prediction

Naive Prediction of Protein Backbone Phi and Psi Dihedral Angles Using Deep Learning

Protein structure prediction represents a significant challenge in the field of bioinformatics, with the prediction of protein structures using backbone dihedral angles recently achieving significant progress due to the rise of deep neural network research. However, there is a trend in protein structure prediction research to employ increasingly complex neural networks and contributions from multiple models. This study, on the other hand, explores how a single model transparently behaves using sequence data only and what can be expected from the predicted angles. To this end, the current paper presents data acquisition, deep learning model definition, and training toward the final protein backbone angle prediction. The method applies a simple fully connected neural network (FCNN) model that takes only the primary structure of the protein with a sliding window of size 21 as input to predict protein backbone ϕ and ψ dihedral angles. Despite its simplicity, the model shows surprising accuracy for the ϕ angle prediction and somewhat lower accuracy for the ψ angle prediction. Moreover, this study demonstrates that protein secondary structure prediction is also possible with simple neural networks that take in only the protein amino-acid residue sequence, but more complex models are required for higher accuracies.

Machine learning approaches demonstrate that protein structures carry information about their genetic coding

Synonymous codons translate into the same amino acid. Although the identity of synonymous codons is often considered inconsequential to the final protein structure, there is mounting evidence for an association between the two. Our study examined this association using regression and classification models, finding that codon sequences predict protein backbone dihedral angles with a lower error than amino acid sequences, and that models trained with true dihedral angles have better classification of synonymous codons given structural information than models trained with random dihedral angles. Using this classification approach, we investigated local codon-codon dependencies and tested whether synonymous codon identity can be predicted more accurately from codon context than amino acid context alone, and most specifically which codon context position carries the most predictive power.

Deep learning for protein secondary structure prediction: Pre and post-AlphaFold

This paper aims to provide a comprehensive review of the trends and challenges of deep neural networks for protein secondary structure prediction (PSSP). In recent years, deep neural networks have become the primary method for protein secondary structure prediction. Previous studies showed that deep neural networks had uplifted the accuracy of three-state secondary structure prediction to more than 80%. Favored deep learning methods, such as convolutional neural networks, recurrent neural networks, inception networks, and graph neural networks, have been implemented in protein secondary structure prediction. Methods adapted from natural language processing (NLP) and computer vision are also employed, including attention mechanism, ResNet, and U-shape networks. In the post-AlphaFold era, PSSP studies focus on different objectives, such as enhancing the quality of evolutionary information and exploiting protein language models as the PSSP input. The recent trend to utilize pre-trained language models as input features for secondary structure prediction provides a new direction for PSSP studies. Moreover, the state-of-the-art accuracy achieved by previous PSSP models is still below its theoretical limit. There are still rooms for improvement to be made in the field.

Machine Learning Based Predictive Model for the Analysis of Sequence Activity Relationships Using Protein Spectra and Protein Descriptors

Accurately establishing the connection between a protein sequence and its function remains a focal point within the field of protein engineering, especially in the context of predicting the effects of mutations. From this, there has been a continued drive to build accurate and reliable predictive models via machine learning that allow for the virtual screening of many protein mutant sequences, measuring the relationship between sequence and 'fitness' or 'activity', commonly known as a Sequence-Activity-Relationship (SAR). An important preliminary stage in the building of these predictive models is the encoding of the chosen sequences. Evaluated in this work is a plethora of encoding strategies using the Amino Acid Index database, where the indices are transformed into their spectral form via Digital Signal Processing (DSP) techniques, as well as numerous protein structural and physiochemical descriptors. The encoding strategies are explored on a dataset curated to measure the thermostability of various mutants from a recombination library, designed from parental cytochrome P450s. In this work it was concluded that the implementation of protein spectra in concatenation with protein descriptors, together with the Partial Least Squares Regression (PLS) algorithm, gave the most noteworthy increase in the quality of the predictive models (as described in Encoding Strategy C), highlighting their utility in identifying an SAR. The accompanying software produced for this paper is termed pySAR (Python Sequence-Activity-Relationship), which allows for a user to find the optimal arrangement of structural and or physiochemical properties to encode their specific mutant library dataset; the source code is available at: https://github.com/amckenna41/pySAR.

Secondary structure specific simpler prediction models for protein backbone angles

Motivation

Protein backbone angle prediction has achieved significant accuracy improvement with the development of deep learning methods. Usually the same deep learning model is used in making prediction for all residues regardless of the categories of secondary structures they belong to. In this paper, we propose to train separate deep learning models for each category of secondary structures. Machine learning methods strive to achieve generality over the training examples and consequently loose accuracy. In this work, we explicitly exploit classification knowledge to restrict generalisation within the specific class of training examples. This is to compensate the loss of generalisation by exploiting specialisation knowledge in an informed way.

Results

The new method named SAP4SS obtains mean absolute error (MAE) values of 15.59, 18.87, 6.03, and 21.71 respectively for four types of backbone angles \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\phi$$\end{document}ϕ, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\psi$$\end{document}ψ, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\theta$$\end{document}θ, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\tau$$\end{document}τ. Consequently, SAP4SS significantly outperforms existing state-of-the-art methods SAP, OPUS-TASS, and SPOT-1D: the differences in MAE for all four types of angles are from 1.5 to 4.1% compared to the best known results.

Availability

SAP4SS along with its data is available from https://gitlab.com/mahnewton/sap4ss.

Accurate prediction of protein torsion angles using evolutionary signatures and recurrent neural network

The amino acid sequence of a protein contains all the necessary information to specify its shape, which dictates its biological activities. However, it is challenging and expensive to experimentally determine the three-dimensional structure of proteins. The backbone torsion angles play a critical role in protein structure prediction, and accurately predicting the angles can considerably advance the tertiary structure prediction by accelerating efficient sampling of the large conformational space for low energy structures. Here we first time propose evolutionary signatures computed from protein sequence profiles, and a novel recurrent architecture, termed ESIDEN, that adopts a straightforward architecture of recurrent neural networks with a small number of learnable parameters. The ESIDEN can capture efficient information from both the classic and new features benefiting from different recurrent architectures in processing information. On the other hand, compared to widely used classic features, the new features, especially the Ramachandran basin potential, provide statistical and evolutionary information to improve prediction accuracy. On four widely used benchmark datasets, the ESIDEN significantly improves the accuracy in predicting the torsion angles by comparison to the best-so-far methods. As demonstrated in the present study, the predicted angles can be used as structural constraints to accurately infer protein tertiary structures. Moreover, the proposed features would pave the way to improve machine learning-based methods in protein folding and structure prediction, as well as function prediction. The source code and data are available at the website https://kornmann.bioch.ox.ac.uk/leri/resources/download.html .

Evolution as a Guide to Designing xeno Amino Acid Alphabets

Here, we summarize a line of remarkably simple, theoretical research to better understand the chemical logic by which life’s standard alphabet of 20 genetically encoded amino acids evolved. The connection to the theme of this Special Issue, “Protein Structure Analysis and Prediction with Statistical Scoring Functions”, emerges from the ways in which current bioinformatics currently lacks empirical science when it comes to xenoproteins composed largely or entirely of amino acids from beyond the standard genetic code. Our intent is to present new perspectives on existing data from two different frontiers in order to suggest fresh ways in which their findings complement one another. These frontiers are origins/astrobiology research into the emergence of the standard amino acid alphabet, and empirical xenoprotein synthesis.

Enhancing protein backbone angle prediction by using simpler models of deep neural networks

Protein structure prediction is a grand challenge. Prediction of protein structures via the representations using backbone dihedral angles has recently achieved significant progress along with the on-going surge of deep neural network (DNN) research in general. However, we observe that in the protein backbone angle prediction research, there is an overall trend to employ more and more complex neural networks and then to throw more and more features to the neural networks. While more features might add more predictive power to the neural network, we argue that redundant features could rather clutter the scenario and more complex neural networks then just could counterbalance the noise. From artificial intelligence and machine learning perspectives, problem representations and solution approaches do mutually interact and thus affect performance. We also argue that comparatively simpler predictors can more easily be reconstructed than the more complex ones. With these arguments in mind, we present a deep learning method named Simpler Angle Predictor (SAP) to train simpler DNN models that enhance protein backbone angle prediction. We then empirically show that SAP significantly outperforms existing state-of-the-art methods on well-known benchmark datasets: for some types of angles, the differences are above 3 in mean absolute error (MAE). The SAP program along with its data is available from the website https://gitlab.com/mahnewton/sap.

SPOT-Fold: Fragment-Free Protein Structure Prediction Guided by Predicted Backbone Structure and Contact Map

Protein structure determination has long been one of the most challenging problems in molecular biology for the past 60 years. Here we present an ab initio protein tertiary-structure prediction method assisted by predicted contact maps from SPOT-Contact and predicted dihedral angles from SPIDER 3. These predicted properties were then fed to the crystallography and NMR system (CNS) for restrained structure modeling. The resulted structures are first evaluated by the potential energy calculated by CNS, followed by dDFIRE energy function for model selections. The method called SPOT-Fold has been tested on 241 CASP targets between 67 and 670 amino acid residues, 60 randomly selected globular proteins under 100 amino acids. The method has a comparable accuracy to other contact-map-based modeling techniques. © 2019 Wiley Periodicals, Inc.

A deep dense inception network for protein beta-turn prediction

Beta-turn prediction is useful in protein function studies and experimental design. Although recent approaches using machine-learning techniques such as support vector machine (SVM), neural networks, and K nearest neighbor have achieved good results for beta-turn prediction, there is still significant room for improvement. As previous predictors utilized features in a sliding window of 4-20 residues to capture interactions among sequentially neighboring residues, such feature engineering may result in incomplete or biased features and neglect interactions among long-range residues. Deep neural networks provide a new opportunity to address these issues. Here, we proposed a deep dense inception network (DeepDIN) for beta-turn prediction, which takes advantage of the state-of-the-art deep neural network design of dense networks and inception networks. A test on a recent BT6376 benchmark data set shows that DeepDIN outperformed the previous best tool BetaTPred3 significantly in both the overall prediction accuracy and the nine-type beta-turn classification accuracy. A tool, called MUFold-BetaTurn, was developed, which is the first beta-turn prediction tool utilizing deep neural networks. The tool can be downloaded at http://dslsrv8.cs.missouri.edu/~cf797/MUFoldBetaTurn/download.html.

End-to-End Differentiable Learning of Protein Structure

Predicting protein structure from sequence is a central challenge of biochemistry. Co-evolution methods show promise, but an explicit sequence-to-structure map remains elusive. Advances in deep learning that replace complex, human-designed pipelines with differentiable models optimized end to end suggest the potential benefits of similarly reformulating structure prediction. Here, we introduce an end-to-end differentiable model for protein structure learning. The model couples local and global protein structure via geometric units that optimize global geometry without violating local covalent chemistry. We test our model using two challenging tasks: predicting novel folds without co-evolutionary data and predicting known folds without structural templates. In the first task, the model achieves state-of-the-art accuracy, and in the second, it comes within 1-2 Å; competing methods using co-evolution and experimental templates have been refined over many years, and it is likely that the differentiable approach has substantial room for further improvement, with applications ranging from drug discovery to protein design.

Improving prediction of protein secondary structure, backbone angles, solvent accessibility and contact numbers by using predicted contact maps and an ensemble of recurrent and residual convolutional neural networks

Motivation

Sequence-based prediction of one dimensional structural properties of proteins has been a long-standing subproblem of protein structure prediction. Recently, prediction accuracy has been significantly improved due to the rapid expansion of protein sequence and structure libraries and advances in deep learning techniques, such as residual convolutional networks (ResNets) and Long-Short-Term Memory Cells in Bidirectional Recurrent Neural Networks (LSTM-BRNNs). Here we leverage an ensemble of LSTM-BRNN and ResNet models, together with predicted residue-residue contact maps, to continue the push towards the attainable limit of prediction for 3- and 8-state secondary structure, backbone angles (θ, τ, ϕ and ψ), half-sphere exposure, contact numbers and solvent accessible surface area (ASA).

Results

The new method, named SPOT-1D, achieves similar, high performance on a large validation set and test set (≈1000 proteins in each set), suggesting robust performance for unseen data. For the large test set, it achieves 87% and 77% in 3- and 8-state secondary structure prediction and 0.82 and 0.86 in correlation coefficients between predicted and measured ASA and contact numbers, respectively. Comparison to current state-of-the-art techniques reveals substantial improvement in secondary structure and backbone angle prediction. In particular, 44% of 40-residue fragment structures constructed from predicted backbone Cα-based θ and τ angles are less than 6 Å root-mean-squared-distance from their native conformations, nearly 20% better than the next best. The method is expected to be useful for advancing protein structure and function prediction.

Availability and implementation

SPOT-1D and its data is available at: http://sparks-lab.org/.

Supplementary information

Supplementary data are available at Bioinformatics online.

Improving Protein Gamma-Turn Prediction Using Inception Capsule Networks

Protein gamma-turn prediction is useful in protein function studies and experimental design. Several methods for gamma-turn prediction have been developed, but the results were unsatisfactory with Matthew correlation coefficients (MCC) around 0.2–0.4. Hence, it is worthwhile exploring new methods for the prediction. A cutting-edge deep neural network, named Capsule Network (CapsuleNet), provides a new opportunity for gamma-turn prediction. Even when the number of input samples is relatively small, the capsules from CapsuleNet are effective to extract high-level features for classification tasks. Here, we propose a deep inception capsule network for gamma-turn prediction. Its performance on the gamma-turn benchmark GT320 achieved an MCC of 0.45, which significantly outperformed the previous best method with an MCC of 0.38. This is the first gamma-turn prediction method utilizing deep neural networks. Also, to our knowledge, it is the first published bioinformatics application utilizing capsule network, which will provide a useful example for the community. Executable and source code can be download at http://dslsrv8.cs.missouri.edu/~cf797/MUFoldGammaTurn/download.html.

Single-sequence-based prediction of protein secondary structures and solvent accessibility by deep whole-sequence learning

Predicting protein structure from sequence alone is challenging. Thus, the majority of methods for protein structure prediction rely on evolutionary information from multiple sequence alignments. In previous work we showed that Long Short-Term Bidirectional Recurrent Neural Networks (LSTM-BRNNs) improved over regular neural networks by better capturing intra-sequence dependencies. Here we show a single-sequence-based prediction method employing LSTM-BRNNs (SPIDER3-Single), that consistently achieves Q3 accuracy of 72.5%, and correlation coefficient of 0.67 between predicted and actual solvent accessible surface area. Moreover, it yields reasonably accurate prediction of eight-state secondary structure, main-chain angles (backbone ϕ and ψ torsion angles and C α-atom-based θ and τ angles), half-sphere exposure, and contact number. The method is more accurate than the corresponding evolutionary-based method for proteins with few sequence homologs, and computationally efficient for large-scale screening of protein-structural properties. It is available as an option in the SPIDER3 server, and a standalone version for download, at http://sparks-lab.org. © 2018 Wiley Periodicals, Inc.

Sixty-five years of the long march in protein secondary structure prediction: the final stretch?

Protein secondary structure prediction began in 1951 when Pauling and Corey predicted helical and sheet conformations for protein polypeptide backbone even before the first protein structure was determined. Sixty-five years later, powerful new methods breathe new life into this field. The highest three-state accuracy without relying on structure templates is now at 82-84%, a number unthinkable just a few years ago. These improvements came from increasingly larger databases of protein sequences and structures for training, the use of template secondary structure information and more powerful deep learning techniques. As we are approaching to the theoretical limit of three-state prediction (88-90%), alternative to secondary structure prediction (prediction of backbone torsion angles and Cα-atom-based angles and torsion angles) not only has more room for further improvement but also allows direct prediction of three-dimensional fragment structures with constantly improved accuracy. About 20% of all 40-residue fragments in a database of 1199 non-redundant proteins have <6 Å root-mean-squared distance from the native conformations by SPIDER2. More powerful deep learning methods with improved capability of capturing long-range interactions begin to emerge as the next generation of techniques for secondary structure prediction. The time has come to finish off the final stretch of the long march towards protein secondary structure prediction.

Prediction of Protein Backbone Torsion Angles Using Deep Residual Inception Neural Networks

Prediction of protein backbone torsion angles (Psi and Phi) can provide important information for protein structure prediction and sequence alignment. Existing methods for Psi-Phi angle prediction have significant room for improvement. In this paper, a new deep residual inception network architecture, called DeepRIN, is proposed for the prediction of Psi-Phi angles. The input to DeepRIN is a feature matrix representing a composition of physico-chemical properties of amino acids, a 20-dimensional position-specific substitution matrix (PSSM) generated by PSI-BLAST, a 30-dimensional hidden Markov Model sequence profile generated by HHBlits, and predicted eight-state secondary structure features. DeepRIN is designed based on inception networks and residual networks that have performed well on image classification and text recognition. The architecture of DeepRIN enables effective encoding of local and global interatcions between amino acids in a protein sequence to achieve accruacte prediction. Extensive experimental results show that DeepRIN outperformed the best existing tools significantly. Compared to the recently released state-of-the-art tool, SPIDER3, DeepRIN reduced the Psi angle prediction error by more than 5 degrees and the Phi angle prediction error by more than 2 degrees on average. The executable tool of DeepRIN is available for download at http://dslsrv8.cs.missouri.edu/~cf797/MUFoldAngle/.

LRFragLib: an effective algorithm to identify fragments for de novo protein structure prediction

Motivation

The quality of fragment library determines the efficiency of fragment assembly, an approach that is widely used in most de novo protein-structure prediction algorithms. Conventional fragment libraries are constructed mainly based on the identities of amino acids, sometimes facilitated by predicted information including dihedral angles and secondary structures. However, it remains challenging to identify near-native fragment structures with low sequence homology.

Results

We introduce a novel fragment-library-construction algorithm, LRFragLib, to improve the detection of near-native low-homology fragments of 7-10 residues, using a multi-stage, flexible selection protocol. Based on logistic regression scoring models, LRFragLib outperforms existing techniques by achieving a significantly higher precision and a comparable coverage on recent CASP protein sets in sampling near-native structures. The method also has a comparable computational efficiency to the fastest existing techniques with substantially reduced memory usage.

Availability and Implementation

The source code is available for download at http://166.111.152.91/Downloads.html.

Contact

hgong@tsinghua.edu.cn.

Supplementary information

Supplementary data are available at Bioinformatics online.

HMMs in Protein Fold Classification

The limitation of most HMMs is their inherent high dimensionality. Therefore we developed several variations of low complexity models that can be applied even to protein families with a few members. In this chapter we present these variations. All of them include the use of a hidden Markov model (HMM), with a small number of states (called reduced state-space HMM), which is trained with both amino acid sequence and secondary structure of proteins whose 3D structure is known and it is used for protein fold classification. We used data from Protein Data Bank and annotation from SCOP database for training and evaluation of the proposed HMM variations for a number of protein folds that belong to major structural classes. Results indicate that the variations have similar performance, or even better in some cases, on classifying proteins than SAM, which is a widely used HMM-based method for protein classification. The major advantage of the proposed variations is that we employed a small number of states and the algorithms used for training and scoring are of low complexity and thus relatively fast. The main variations examined include a version of the reduced state-space HMM with seven states (7-HMM), a version of the reduced state-space HMM with three states (3-HMM) and an optimized version of the reduced state-space HMM with three states, where an optimization process is applied to its scores (optimized 3-HMM).

Improving prediction of secondary structure, local backbone angles, and solvent accessible surface area of proteins by iterative deep learning

Direct prediction of protein structure from sequence is a challenging problem. An effective approach is to break it up into independent sub-problems. These sub-problems such as prediction of protein secondary structure can then be solved independently. In a previous study, we found that an iterative use of predicted secondary structure and backbone torsion angles can further improve secondary structure and torsion angle prediction. In this study, we expand the iterative features to include solvent accessible surface area and backbone angles and dihedrals based on Cα atoms. By using a deep learning neural network in three iterations, we achieved 82% accuracy for secondary structure prediction, 0.76 for the correlation coefficient between predicted and actual solvent accessible surface area, 19° and 30° for mean absolute errors of backbone φ and ψ angles, respectively, and 8° and 32° for mean absolute errors of Cα-based θ and τ angles, respectively, for an independent test dataset of 1199 proteins. The accuracy of the method is slightly lower for 72 CASP 11 targets but much higher than those of model structures from current state-of-the-art techniques. This suggests the potentially beneficial use of these predicted properties for model assessment and ranking.

Predicting backbone Cα angles and dihedrals from protein sequences by stacked sparse auto-encoder deep neural network

Because a nearly constant distance between two neighbouring Cα atoms, local backbone structure of proteins can be represented accurately by the angle between C(αi-1)-C(αi)-C(αi+1) (θ) and a dihedral angle rotated about the C(αi)-C(αi+1) bond (τ). θ and τ angles, as the representative of structural properties of three to four amino-acid residues, offer a description of backbone conformations that is complementary to φ and ψ angles (single residue) and secondary structures (>3 residues). Here, we report the first machine-learning technique for sequence-based prediction of θ and τ angles. Predicted angles based on an independent test have a mean absolute error of 9° for θ and 34° for τ with a distribution on the θ-τ plane close to that of native values. The average root-mean-square distance of 10-residue fragment structures constructed from predicted θ and τ angles is only 1.9Å from their corresponding native structures. Predicted θ and τ angles are expected to be complementary to predicted ϕ and ψ angles and secondary structures for using in model validation and template-based as well as template-free structure prediction. The deep neural network learning technique is available as an on-line server called Structural Property prediction with Integrated DEep neuRal network (SPIDER) at http://sparks-lab.org.

Assessment of optimized Markov models in protein fold classification

Protein fold classification is a challenging task strongly associated with the determination of proteins' structure. In this work, we tested an optimization strategy on a Markov chain and a recently introduced Hidden Markov Model (HMM) with reduced state-space topology. The proteins with unknown structure were scored against both these models. Then the derived scores were optimized following a local optimization method. The Protein Data Bank (PDB) and the annotation of the Structural Classification of Proteins (SCOP) database were used for the evaluation of the proposed methodology. The results demonstrated that the fold classification accuracy of the optimized HMM was substantially higher compared to that of the Markov chain or the reduced state-space HMM approaches. The proposed methodology achieved an accuracy of 41.4% on fold classification, while Sequence Alignment and Modeling (SAM), which was used for comparison, reached an accuracy of 38%.

Prediction and validation of the unexplored RNA-binding protein atlas of the human proteome

Detecting protein-RNA interactions is challenging both experimentally and computationally because RNAs are large in number, diverse in cellular location and function, and flexible in structure. As a result, many RNA-binding proteins (RBPs) remain to be identified. Here, a template-based, function-prediction technique SPOT-Seq for RBPs is applied to human proteome and its result is validated by a recent proteomic experimental discovery of 860 mRNA-binding proteins (mRBPs). The coverage (or sensitivity) is 42.6% for 1217 known RBPs annotated in the Gene Ontology and 43.6% for 860 newly discovered human mRBPs. Consistent sensitivity indicates the robust performance of SPOT-Seq for predicting RBPs. More importantly, SPOT-Seq detects 2418 novel RBPs in human proteome, 291 of which were validated by the newly discovered mRBP set. Among 291 validated novel RBPs, 61 are not homologous to any known RBPs. Successful validation of predicted novel RBPs permits us to further analysis of their phenotypic roles in disease pathways. The dataset of 2418 predicted novel RBPs along with confidence levels and complex structures is available at http://sparks-lab.org (in publications) for experimental confirmations and hypothesis generation.

Perspective: Coarse-grained models for biomolecular systems

By focusing on essential features, while averaging over less important details, coarse-grained (CG) models provide significant computational and conceptual advantages with respect to more detailed models. Consequently, despite dramatic advances in computational methodologies and resources, CG models enjoy surging popularity and are becoming increasingly equal partners to atomically detailed models. This perspective surveys the rapidly developing landscape of CG models for biomolecular systems. In particular, this review seeks to provide a balanced, coherent, and unified presentation of several distinct approaches for developing CG models, including top-down, network-based, native-centric, knowledge-based, and bottom-up modeling strategies. The review summarizes their basic philosophies, theoretical foundations, typical applications, and recent developments. Additionally, the review identifies fundamental inter-relationships among the diverse approaches and discusses outstanding challenges in the field. When carefully applied and assessed, current CG models provide highly efficient means for investigating the biological consequences of basic physicochemical principles. Moreover, rigorous bottom-up approaches hold great promise for further improving the accuracy and scope of CG models for biomolecular systems.

LEAP: highly accurate prediction of protein loop conformations by integrating coarse-grained sampling and optimized energy scores with all-atom refinement of backbone and side chains

Prediction of protein loop conformations without any prior knowledge (ab initio prediction) is an unsolved problem. Its solution will significantly impact protein homology and template-based modeling as well as ab initio protein-structure prediction. Here, we developed a coarse-grained, optimized scoring function for initial sampling and ranking of loop decoys. The resulting decoys are then further optimized in backbone and side-chain conformations and ranked by all-atom energy scoring functions. The final integrated technique called loop prediction by energy-assisted protocol achieved a median value of 2.1 Å root mean square deviation (RMSD) for 325 12-residue test loops and 2.0 Å RMSD for 45 12-residue loops from critical assessment of structure-prediction techniques (CASP) 10 target proteins with native core structures (backbone and side chains). If all side-chain conformations in protein cores were predicted in the absence of the target loop, loop-prediction accuracy only reduces slightly (0.2 Å difference in RMSD for 12-residue loops in the CASP target proteins). The accuracy obtained is about 1 Å RMSD or more improvement over other methods we tested. The executable file for a Linux system is freely available for academic users at http://sparks-lab.org.

SPOT-Seq-RNA: predicting protein-RNA complex structure and RNA-binding function by fold recognition and binding affinity prediction

RNA-binding proteins (RBPs) play key roles in RNA metabolism and post-transcriptional regulation. Computational methods have been developed separately for prediction of RBPs and RNA-binding residues by machine-learning techniques and prediction of protein-RNA complex structures by rigid or semiflexible structure-to-structure docking. Here, we describe a template-based technique called SPOT-Seq-RNA that integrates prediction of RBPs, RNA-binding residues, and protein-RNA complex structures into a single package. This integration is achieved by combining template-based structure-prediction software, SPARKS X, with binding affinity prediction software, DRNA. This tool yields reasonable sensitivity (46 %) and high precision (84 %) for an independent test set of 215 RBPs and 5,766 non-RBPs. SPOT-Seq-RNA is computationally efficient for genome-scale prediction of RBPs and protein-RNA complex structures. Its application to human genome study has revealed a similar sensitivity and ability to uncover hundreds of novel RBPs beyond simple homology. The online server and downloadable version of SPOT-Seq-RNA are available at http://sparks-lab.org/server/SPOT-Seq-RNA/.

The intrinsic conformational features of amino acids from a protein coil library and their applications in force field development

The local conformational (φ, ψ, χ) preferences of amino acid residues remain an active research area, which are important for the development of protein force fields. In this perspective article, we first summarize spectroscopic studies of alanine-based short peptides in aqueous solution. While most studies indicate a preference for the P(II) conformation in the unfolded state over α and β conformations, significant variations are also observed. A statistical analysis from various coil libraries of high-resolution protein structures is then summarized, which gives a more coherent view of the local conformational features. The φ, ψ, χ distributions of the 20 amino acids have been obtained from a protein coil library, considering both backbone and side-chain conformational preferences. The intrinsic side-chain χ(1) rotamer preference and χ(1)-dependent Ramachandran plot can be generally understood by combining the interaction of the side-chain Cγ/Oγ atom with two neighboring backbone peptide groups. Current all-atom force fields such as AMBER ff99sb-ILDN, ff03 and OPLS-AA/L do not reproduce these distributions well. A method has been developed by combining the φ, ψ plot of alanine with the influence of side-chain χ(1) rotamers to derive the local conformational features of various amino acids. It has been further applied to improve the OPLS-AA force field. The modified force field (OPLS-AA/C) reproduces experimental (3)J coupling constants for various short peptides quite well. It also better reproduces the temperature-dependence of the helix-coil transition for alanine-based peptides. The new force field can fold a series of peptides and proteins with various secondary structures to their experimental structures. MD simulations of several globular proteins using the improved force field give significantly less deviation (RMSD) to experimental structures. The results indicate that the local conformational features from coil libraries are valuable for the development of balanced protein force fields.

Relative free enthalpies for point mutations in two proteins with highly similar sequences but different folds

Enveloping distribution sampling was used to calculate free-enthalpy changes associated with single amino acid mutations for a pair of proteins, GA95 and GB95, that show 95% sequence identity yet fold into topologically different structures. Of the L → A, I → F, and L → Y mutations at positions 20, 30, and 45, respectively, of the 56-residue sequence, the first and the last contribute the most to the free-enthalpy difference between the native and non-native sequence-structure combinations, in agreement with the experimental findings for this protein pair. The individual free-enthalpy changes are almost sequence-independent in the four-strand/one-helix structure, the stable form of GB95, while in the three-helix bundle structure, the stable form of GA95, an interplay between residues 20 and 45 is observed.

Assessing protein conformational sampling methods based on bivariate lag-distributions of backbone angles

Despite considerable progress in the past decades, protein structure prediction remains one of the major unsolved problems in computational biology. Angular-sampling-based methods have been extensively studied recently due to their ability to capture the continuous conformational space of protein structures. The literature has focused on using a variety of parametric models of the sequential dependencies between angle pairs along the protein chains. In this article, we present a thorough review of angular-sampling-based methods by assessing three main questions: What is the best distribution type to model the protein angles? What is a reasonable number of components in a mixture model that should be considered to accurately parameterize the joint distribution of the angles? and What is the order of the local sequence-structure dependency that should be considered by a prediction method? We assess the model fits for different methods using bivariate lag-distributions of the dihedral/planar angles. Moreover, the main information across the lags can be extracted using a technique called Lag singular value decomposition (LagSVD), which considers the joint distribution of the dihedral/planar angles over different lags using a nonparametric approach and monitors the behavior of the lag-distribution of the angles using singular value decomposition. As a result, we developed graphical tools and numerical measurements to compare and evaluate the performance of different model fits. Furthermore, we developed a web-tool (http://www.stat.tamu.edu/∼madoliat/LagSVD) that can be used to produce informative animations.

Comprehensive conformational studies of five tripeptides and a deduced method for efficient determinations of peptide structures

Thorough searches on the potential energy surfaces of five tripeptides, GGG, GYG, GWG, TGG, and MGG, were performed by considering all possible combinations of the bond rotational degrees of freedom with a semiempirical and ab initio combined computational approach. Structural characteristics of the obtained stable tripeptide conformers were carefully analyzed. Conformers of the five tripeptides were found to be closely connected with conformers of their constituting dipeptides and amino acids. A method for finding all important tripeptide conformers by optimizing a small number of trial structures generated by suitable superposition of the parent amino acid and dipeptide conformers is thus proposed. Applying the method to another five tripeptides, YGG, FGG, WGG, GFA, and GGF, studied before shows that the new approach is both efficient and reliable by providing the most complete ensembles of tripeptide conformers. The method is further generalized for application to larger peptides by introducing the breeding and mutation concepts in a genetic algorithm way. The generalized method is verified to be capable of finding tetrapeptide conformers with secondary structures of strands, helices, and turns, which are highly populated in larger peptides. This show some promise for the proposed method to be applied for the structural determination of larger peptides.

Improving protein fold recognition and template-based modeling by employing probabilistic-based matching between predicted one-dimensional structural properties of query and corresponding native properties of templates

MOTIVATION

In recent years, development of a single-method fold-recognition server lags behind consensus and multiple template techniques. However, a good consensus prediction relies on the accuracy of individual methods. This article reports our efforts to further improve a single-method fold recognition technique called SPARKS by changing the alignment scoring function and incorporating the SPINE-X techniques that make improved prediction of secondary structure, backbone torsion angle and solvent accessible surface area.

RESULTS

The new method called SPARKS-X was tested with the SALIGN benchmark for alignment accuracy, Lindahl and SCOP benchmarks for fold recognition, and CASP 9 blind test for structure prediction. The method is compared to several state-of-the-art techniques such as HHPRED and BoostThreader. Results show that SPARKS-X is one of the best single-method fold recognition techniques. We further note that incorporating multiple templates and refinement in model building will likely further improve SPARKS-X.

AVAILABILITY

The method is available as a SPARKS-X server at http://sparks.informatics.iupui.edu/