ANGLOR: a composite machine-learning algorithm for protein backbone torsion angle 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.

TAFPred: Torsion Angle Fluctuations Prediction from Protein Sequences

Protein molecules show varying degrees of flexibility throughout their three-dimensional structures. The flexibility is determined by the fluctuations in torsion angles, specifically phi (φ) and psi (ψ), which define the protein backbone. These angle fluctuations are derived from variations in backbone torsion angles observed in different models. By analyzing the fluctuations in Cartesian coordinate space, we can understand the structural flexibility of proteins. Predicting torsion angle fluctuations is valuable for determining protein function and structure when these angles act as constraints. In this study, a machine learning method called TAFPred is developed to predict torsion angle fluctuations using protein sequences directly. The method incorporates various features, such as disorder probability, position-specific scoring matrix profiles, secondary structure probabilities, and more. TAFPred, employing an optimized Light Gradient Boosting Machine Regressor (LightGBM), achieved high accuracy with correlation coefficients of 0.746 and 0.737 and mean absolute errors of 0.114 and 0.123 for the φ and ψ angles, respectively. Compared to the state-of-the-art method, TAFPred demonstrated significant improvements of 10.08% in MAE and 24.83% in PCC for the phi angle and 9.93% in MAE, and 22.37% in PCC for the psi angle.

IGPRED-MultiTask: A Deep Learning Model to Predict Protein Secondary Structure, Torsion Angles and Solvent Accessibility

Protein secondary structure, solvent accessibility and torsion angle predictions are preliminary steps to predict 3D structure of a protein. Deep learning approaches have achieved significant improvements in predicting various features of protein structure. In this study, IGPRED-Multitask, a deep learning model with multi task learning architecture based on deep inception network, graph convolutional network and a bidirectional long short-term memory is proposed. Moreover, hyper-parameters of the model are fine-tuned using Bayesian optimization, which is faster and more effective than grid search. The same benchmark test data sets as in the OPUS-TASS paper including TEST2016, TEST2018, CASP12, CASP13, CASPFM, HARD68, CAMEO93, CAMEO93_HARD, as well as the train and validation sets, are used for fair comparison with the literature. Statistically significant improvements are observed in secondary structure prediction on 4 datasets, in phi angle prediction on 2 datasets and in psi angel prediction on 3 datasets compared to the state-of-the-art methods. For solvent accessibility prediction, TEST2016 and TEST2018 datasets are used only to assess the performance of the proposed model.

Personal Precise Force Field for Intrinsically Disordered and Ordered Proteins Based on Deep Learning

Intrinsically disordered proteins (IDPs) are proteins without a fixed three-dimensional (3D) structure under physiological conditions and are associated with Parkinson's disease, Alzheimer's disease, cancer, cardiovascular disease, amyloidosis, diabetes, and other diseases. Experimental methods can hardly capture the ensemble of diverse conformations for IDPs. Molecular dynamics (MD) simulations can sample continuous conformations that might provide a valuable complement to experimental data. However, the accuracy of MD simulations depends on the quality of force field. In particular, the evolutionary conservation and coevolution of IDPs introduce that current force fields could not precisely reproduce the conformation of IDPs. In order to improve the performance of force field, deep learning and reweighting methods were used to automatically generate personal force field parameters for intrinsically disordered and ordered proteins. At first, the deep learning method predicted more accuracy φ/ψ dihedral of residue than the previous method. Then, reweighting optimized the personal force field parameters for each residue. Finally, typical representative systems such as IDPs, structure protein, and fast-folding protein were used to evaluate this force field. The results indicate that two personal force field parameters (named PPFF1 and PPFF1_af2) could better reproduce the experimental observables than ff03CMAP force field. In summary, this strategy will provide feasibility for the development of precise personal force fields.

Prediction of metal ion ligand binding residues by adding disorder value and propensity factors based on deep learning algorithm

Proteins need to interact with different ligands to perform their functions. Among the ligands, the metal ion is a major ligand. At present, the prediction of protein metal ion ligand binding residues is a challenge. In this study, we selected Zn²⁺, Cu²⁺, Fe²⁺, Fe³⁺, Co²⁺, Mn²⁺, Ca²⁺ and Mg²⁺ metal ion ligands from the BioLip database as the research objects. Based on the amino acids, the physicochemical properties and predicted structural information, we introduced the disorder value as the feature parameter. In addition, based on the component information, position weight matrix and information entropy, we introduced the propensity factor as prediction parameters. Then, we used the deep neural network algorithm for the prediction. Furtherly, we made an optimization for the hyper-parameters of the deep learning algorithm and obtained improved results than the previous IonSeq method.

cpxDeepMSA: A Deep Cascade Algorithm for Constructing Multiple Sequence Alignments of Protein–Protein Interactions

Protein–protein interactions (PPIs) are fundamental to many biological processes. The coevolution-based prediction of interacting residues has made great strides in protein complexes that are known to interact. A multiple sequence alignment (MSA) is the basis of coevolution analysis. MSAs have recently made significant progress in the protein monomer sequence analysis. However, no standard or efficient pipelines are available for the sensitive protein complex MSA (cpxMSA) collection. How to generate cpxMSA is one of the most challenging problems of sequence coevolution analysis. Although several methods have been developed to address this problem, no standalone program exists. Furthermore, the number of built-in properties is limited; hence, it is often difficult for users to analyze sequence coevolution according to their desired cpxMSA. In this article, we developed a novel cpxMSA approach (cpxDeepMSA. We used different protein monomer databases and incorporated the three strategies (genomic distance, phylogeny information, and STRING interaction network) used to join the monomer MSA results of protein complexes, which can prevent using a single method fail to the joint two-monomer MSA causing the cpxMSA construction failure. We anticipate that the cpxDeepMSA algorithm will become a useful high-throughput tool in protein complex structure predictions, inter-protein residue-residue contacts, and the biological sequence coevolution analysis.

Application of the MAHDS Method for Multiple Alignment of Highly Diverged Amino Acid Sequences

The aim of this work was to compare the multiple alignment methods MAHDS, T-Coffee, MUSCLE, Clustal Omega, Kalign, MAFFT, and PRANK in their ability to align highly divergent amino acid sequences. To accomplish this, we created test amino acid sequences with an average number of substitutions per amino acid (x) from 0.6 to 5.6, a total of 81 sets. Comparison of the performance of sequence alignments constructed by MAHDS and previously developed algorithms using the CS and Z score criteria and the benchmark alignment database (BAliBASE) indicated that, although the quality of the alignments built with MAHDS was somewhat lower than that of the other algorithms, it was compensated by greater statistical significance. MAHDS could construct statistically significant alignments of artificial sequences with x ≤ 4.8, whereas the other algorithms (T-Coffee, MUSCLE, Clustal Omega, Kalign, MAFFT, and PRANK) could not perform that at x > 2.4. The application of MAHDS to align 21 families of highly diverged proteins (identity < 20%) from Pfam and HOMSTRAD databases showed that it could calculate statistically significant alignments in cases when the other methods failed. Thus, MAHDS could be used to construct statistically significant multiple alignments of highly divergent protein sequences, which accumulated multiple mutations during evolution.

Recognition of Metal Ion Ligand-Binding Residues by Adding Correlation Features and Propensity Factors

The realization of many protein functions is inseparable from the interaction with ligands; in particular, the combination of protein and metal ion ligands performs an important biological function. Currently, it is a challenging work to identify the metal ion ligand-binding residues accurately by computational approaches. In this study, we proposed an improved method to predict the binding residues of 10 metal ion ligands (Zn²⁺, Cu²⁺, Fe²⁺, Fe³⁺, Co²⁺, Mn²⁺, Ca²⁺, Mg²⁺, Na⁺, and K⁺). Based on the basic feature parameters of amino acids, and physicochemical and predicted structural information, we added another two features of amino acid correlation information and binding residue propensity factors. With the optimized parameters, we used the GBM algorithm to predict metal ion ligand-binding residues. In the obtained results, the Sn and MCC values were over 10.17% and 0.297, respectively. Besides, the S_n and MCC values of transition metals were higher than 34.46% and 0.564, respectively. In order to test the validity of our model, another method (Random Forest) was also used in comparison. The better results of this work indicated that the proposed method would be a valuable tool to predict metal ion ligand-binding residues.

Predicting Ca2+ and Mg2+ ligand binding sites by deep neural network algorithm

Background

Alkaline earth metal ions are important protein binding ligands in human body, and it is of great significance to predict their binding residues.

Results

In this paper, Mg²⁺ and Ca²⁺ ligands are taken as the research objects. Based on the characteristic parameters of protein sequences, amino acids, physicochemical characteristics of amino acids and predicted structural information, deep neural network algorithm is used to predict the binding sites of proteins. By optimizing the hyper-parameters of the deep learning algorithm, the prediction results by the fivefold cross-validation are better than those of the Ionseq method. In addition, to further verify the performance of the proposed model, the undersampling data processing method is adopted, and the prediction results on independent test are better than those obtained by the support vector machine algorithm.

Conclusions

An efficient method for predicting Mg²⁺ and Ca²⁺ ligand binding sites was presented.

Artificial intelligence in the prediction of protein-ligand interactions: recent advances and future directions

New drug production, from target identification to marketing approval, takes over 12 years and can cost around $2.6 billion. Furthermore, the COVID-19 pandemic has unveiled the urgent need for more powerful computational methods for drug discovery. Here, we review the computational approaches to predicting protein-ligand interactions in the context of drug discovery, focusing on methods using artificial intelligence (AI). We begin with a brief introduction to proteins (targets), ligands (e.g. drugs) and their interactions for nonexperts. Next, we review databases that are commonly used in the domain of protein-ligand interactions. Finally, we survey and analyze the machine learning (ML) approaches implemented to predict protein-ligand binding sites, ligand-binding affinity and binding pose (conformation) including both classical ML algorithms and recent deep learning methods. After exploring the correlation between these three aspects of protein-ligand interaction, it has been proposed that they should be studied in unison. We anticipate that our review will aid exploration and development of more accurate ML-based prediction strategies for studying protein-ligand interactions.

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 .

AI applications in functional genomics

We review the current applications of artificial intelligence (AI) in functional genomics. The recent explosion of AI follows the remarkable achievements made possible by "deep learning", along with a burst of "big data" that can meet its hunger. Biology is about to overthrow astronomy as the paradigmatic representative of big data producer. This has been made possible by huge advancements in the field of high throughput technologies, applied to determine how the individual components of a biological system work together to accomplish different processes. The disciplines contributing to this bulk of data are collectively known as functional genomics. They consist in studies of: i) the information contained in the DNA (genomics); ii) the modifications that DNA can reversibly undergo (epigenomics); iii) the RNA transcripts originated by a genome (transcriptomics); iv) the ensemble of chemical modifications decorating different types of RNA transcripts (epitranscriptomics); v) the products of protein-coding transcripts (proteomics); and vi) the small molecules produced from cell metabolism (metabolomics) present in an organism or system at a given time, in physiological or pathological conditions. After reviewing main applications of AI in functional genomics, we discuss important accompanying issues, including ethical, legal and economic issues and the importance of explainability.

Toward the solution of the protein structure prediction problem

Since Anfinsen demonstrated that the information encoded in a protein's amino acid sequence determines its structure in 1973, solving the protein structure prediction problem has been the Holy Grail of structural biology. The goal of protein structure prediction approaches is to utilize computational modeling to determine the spatial location of every atom in a protein molecule starting from only its amino acid sequence. Depending on whether homologous structures can be found in the Protein Data Bank (PDB), structure prediction methods have been historically categorized as template-based modeling (TBM) or template-free modeling (FM) approaches. Until recently, TBM has been the most reliable approach to predicting protein structures, and in the absence of reliable templates, the modeling accuracy sharply declines. Nevertheless, the results of the most recent community-wide assessment of protein structure prediction experiment (CASP14) have demonstrated that the protein structure prediction problem can be largely solved through the use of end-to-end deep machine learning techniques, where correct folds could be built for nearly all single-domain proteins without using the PDB templates. Critically, the model quality exhibited little correlation with the quality of available template structures, as well as the number of sequence homologs detected for a given target protein. Thus, the implementation of deep-learning techniques has essentially broken through the 50-year-old modeling border between TBM and FM approaches and has made the success of high-resolution structure prediction significantly less dependent on template availability in the PDB library.

DeepMSA: constructing deep multiple sequence alignment to improve contact prediction and fold-recognition for distant-homology proteins

MOTIVATION

The success of genome sequencing techniques has resulted in rapid explosion of protein sequences. Collections of multiple homologous sequences can provide critical information to the modeling of structure and function of unknown proteins. There are however no standard and efficient pipeline available for sensitive multiple sequence alignment (MSA) collection. This is particularly challenging when large whole-genome and metagenome databases are involved.

RESULTS

We developed DeepMSA, a new open-source method for sensitive MSA construction, which has homologous sequences and alignments created from multi-sources of whole-genome and metagenome databases through complementary hidden Markov model algorithms. The practical usefulness of the pipeline was examined in three large-scale benchmark experiments based on 614 non-redundant proteins. First, DeepMSA was utilized to generate MSAs for residue-level contact prediction by six coevolution and deep learning-based programs, which resulted in an accuracy increase in long-range contacts by up to 24.4% compared to the default programs. Next, multiple threading programs are performed for homologous structure identification, where the average TM-score of the template alignments has over 7.5% increases with the use of the new DeepMSA profiles. Finally, DeepMSA was used for secondary structure prediction and resulted in statistically significant improvements in the Q3 accuracy. It is noted that all these improvements were achieved without re-training the parameters and neural-network models, demonstrating the robustness and general usefulness of the DeepMSA in protein structural bioinformatics applications, especially for targets without homologous templates in the PDB library.

AVAILABILITY AND IMPLEMENTATION

https://zhanglab.ccmb.med.umich.edu/DeepMSA/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

OPUS-TASS: a protein backbone torsion angles and secondary structure predictor based on ensemble neural networks

Motivation

Predictions of protein backbone torsion angles (ϕ and ψ) and secondary structure from sequence are crucial subproblems in protein structure prediction. With the development of deep learning approaches, their accuracies have been significantly improved. To capture the long-range interactions, most studies integrate bidirectional recurrent neural networks into their models. In this study, we introduce and modify a recently proposed architecture named Transformer to capture the interactions between the two residues theoretically with arbitrary distance. Moreover, we take advantage of multitask learning to improve the generalization of neural network by introducing related tasks into the training process. Similar to many previous studies, OPUS-TASS uses an ensemble of models and achieves better results.

Results

OPUS-TASS uses the same training and validation sets as SPOT-1D. We compare the performance of OPUS-TASS and SPOT-1D on TEST2016 (1213 proteins) and TEST2018 (250 proteins) proposed in the SPOT-1D paper, CASP12 (55 proteins), CASP13 (32 proteins) and CASP-FM (56 proteins) proposed in the SAINT paper, and a recently released PDB structure collection from CAMEO (93 proteins) named as CAMEO93. On these six test sets, OPUS-TASS achieves consistent improvements in both backbone torsion angles prediction and secondary structure prediction. On CAMEO93, SPOT-1D achieves the mean absolute errors of 16.89 and 23.02 for ϕ and ψ predictions, respectively, and the accuracies for 3- and 8-state secondary structure predictions are 87.72 and 77.15%, respectively. In comparison, OPUS-TASS achieves 16.56 and 22.56 for ϕ and ψ predictions, and 89.06 and 78.87% for 3- and 8-state secondary structure predictions, respectively. In particular, after using our torsion angles refinement method OPUS-Refine as the post-processing procedure for OPUS-TASS, the mean absolute errors for final ϕ and ψ predictions are further decreased to 16.28 and 21.98, respectively.

Availability and implementation

The training and the inference codes of OPUS-TASS and its data are available at https://github.com/thuxugang/opus_tass.

Supplementary information

Supplementary data are available at Bioinformatics online.

Recognizing Ion Ligand-Binding Residues by Random Forest Algorithm Based on Optimized Dihedral Angle

The prediction of ion ligand–binding residues in protein sequences is a challenging work that contributes to understand the specific functions of proteins in life processes. In this article, we selected binding residues of 14 ion ligands as research objects, including four acid radical ion ligands and 10 metal ion ligands. Based on the amino acid sequence information, we selected the composition and position conservation information of amino acids, the predicted structural information, and physicochemical properties of amino acids as basic feature parameters. We then performed a statistical analysis and reclassification for dihedral angle and proposed new methods on the extraction of feature parameters. The methods mainly included applying information entropy on the extraction of polarization charge and hydrophilic–hydrophobic information of amino acids and using position weight matrices on the extraction of position conservation information. In the prediction model, we used the random forest algorithm and obtained better prediction results than previous works. With the independent test, the Matthew's correlation coefficient and accuracy of 10 metal ion ligand–binding residues were larger than 0.07 and 52%, respectively; the corresponding evaluation values of four acid radical ion ligand–binding residues were larger than 0.15 and 86%, respectively. Further, we classified and combined the phi and psi angles and optimized prediction model for each ion ligand–binding residue.

AllesTM: predicting multiple structural features of transmembrane proteins

Background

This study is motivated by the following three considerations: a) the physico-chemical properties of transmembrane (TM) proteins are distinctly different from those of globular proteins, necessitating the development of specialized structure prediction techniques, b) for many structural features no specialized predictors for TM proteins are available at all, and c) deep learning algorithms allow to automate the feature engineering process and thus facilitate the development of multi-target methods for predicting several protein properties at once.

Results

We present AllesTM, an integrated tool to predict almost all structural features of transmembrane proteins that can be extracted from atomic coordinate data. It blends several machine learning algorithms: random forests and gradient boosting machines, convolutional neural networks in their original form as well as those enhanced by dilated convolutions and residual connections, and, finally, long short-term memory architectures. AllesTM outperforms other available methods in predicting residue depth in the membrane, flexibility, topology, relative solvent accessibility in its bound state, while in torsion angles, secondary structure and monomer relative solvent accessibility prediction it lags only slightly behind the currently leading technique SPOT-1D. High accuracy on a multitude of prediction targets and easy installation make AllesTM a one-stop shop for many typical problems in the structural bioinformatics of transmembrane proteins.

Conclusions

In addition to presenting a highly accurate prediction method and eliminating the need to install and maintain many different software tools, we also provide a comprehensive overview of the impact of different machine learning algorithms and parameter choices on the prediction performance.

AllesTM is freely available at https://github.com/phngs/allestm.

The Identification of Metal Ion Ligand-Binding Residues by Adding the Reclassified Relative Solvent Accessibility

Many proteins realize their special functions by binding with specific metal ion ligands during a cell's life cycle. The ability to correctly identify metal ion ligand-binding residues is valuable for the human health and the design of molecular drug. Precisely identifying these residues, however, remains challenging work. We have presented an improved computational approach for predicting the binding residues of 10 metal ion ligands (Zn2+, Cu2+, Fe2+, Fe3+, Co2+, Ca2+, Mg2+, Mn2+, Na+, and K+) by adding reclassified relative solvent accessibility (RSA). The best accuracy of fivefold cross-validation was higher than 77.9%, which was about 16% higher than the previous result on the same dataset. It was found that different reclassification of the RSA information can make different contributions to the identification of specific ligand binding residues. Our study has provided an additional understanding of the effect of the RSA on the identification of metal ion ligand binding residues.

Recognizing ion ligand binding sites by SMO algorithm

Background

In many important life activities, the execution of protein function depends on the interaction between proteins and ligands. As an important protein binding ligand, the identification of the binding site of the ion ligands plays an important role in the study of the protein function.

Results

In this study, four acid radical ion ligands (NO₂⁻,CO₃²⁻,SO₄²⁻,PO₄³⁻) and ten metal ion ligands (Zn²⁺,Cu²⁺,Fe²⁺,Fe³⁺,Ca²⁺,Mg²⁺,Mn²⁺,Na⁺,K⁺,Co²⁺) are selected as the research object, and the Sequential minimal optimization (SMO) algorithm based on sequence information was proposed, better prediction results were obtained by 5-fold cross validation.

Conclusions

An efficient method for predicting ion ligand binding sites was presented.

Investigation of machine learning techniques on proteomics: A comprehensive survey

Proteomics is the extensive investigation of proteins which has empowered the recognizable proof of consistently expanding quantities of protein. Proteins are necessary part of living life form, with numerous capacities. The proteome is the complete arrangement of proteins that are created or altered by a life form or framework of the organism. Proteome fluctuates with time and unambiguous prerequisites, or stresses, that a cell or organism experiences. Proteomics is an interdisciplinary area that has derived from the hereditary data of different genome ventures. Much proteomics information is gathered with the assistance of high throughput techniques, for example, mass spectrometry and microarray. It would regularly take weeks or months to analyze the information and perform examinations by hand. Therefore, scholars and scientific experts are teaming up with computer science researchers and mathematicians to make projects and pipeline to computationally examine the protein information. Utilizing bioinformatics procedures, scientists are prepared to do quicker investigation and protein information storing. The goal of this paper is to brief about the review of machine learning procedures and its application in the field of proteomics.

Predicting protein-ligand binding residues with deep convolutional neural networks

Background

Ligand-binding proteins play key roles in many biological processes. Identification of protein-ligand binding residues is important in understanding the biological functions of proteins. Existing computational methods can be roughly categorized as sequence-based or 3D-structure-based methods. All these methods are based on traditional machine learning. In a series of binding residue prediction tasks, 3D-structure-based methods are widely superior to sequence-based methods. However, due to the great number of proteins with known amino acid sequences, sequence-based methods have considerable room for improvement with the development of deep learning. Therefore, prediction of protein-ligand binding residues with deep learning requires study.

Results

In this study, we propose a new sequence-based approach called DeepCSeqSite for ab initio protein-ligand binding residue prediction. DeepCSeqSite includes a standard edition and an enhanced edition. The classifier of DeepCSeqSite is based on a deep convolutional neural network. Several convolutional layers are stacked on top of each other to extract hierarchical features. The size of the effective context scope is expanded as the number of convolutional layers increases. The long-distance dependencies between residues can be captured by the large effective context scope, and stacking several layers enables the maximum length of dependencies to be precisely controlled. The extracted features are ultimately combined through one-by-one convolution kernels and softmax to predict whether the residues are binding residues. The state-of-the-art ligand-binding method COACH and some of its submethods are selected as baselines. The methods are tested on a set of 151 nonredundant proteins and three extended test sets. Experiments show that the improvement of the Matthews correlation coefficient (MCC) is no less than 0.05. In addition, a training data augmentation method that slightly improves the performance is discussed in this study.

Conclusions

Without using any templates that include 3D-structure data, DeepCSeqSite significantlyoutperforms existing sequence-based and 3D-structure-based methods, including COACH. Augmentation of the training sets slightly improves the performance. The model, code and datasets are available at https://github.com/yfCuiFaith/DeepCSeqSite.

ccPDB 2.0: an updated version of datasets created and compiled from Protein Data Bank

ccPDB 2.0 (http://webs.iiitd.edu.in/raghava/ccpdb) is an updated version of the manually curated database ccPDB that maintains datasets required for developing methods to predict the structure and function of proteins. The number of datasets compiled from literature increased from 45 to 141 in ccPDB 2.0. Similarly, the number of protein structures used for creating datasets also increased from ~74 000 to ~137 000 (PDB March 2018 release). ccPDB 2.0 provides the same web services and flexible tools which were present in the previous version of the database. In the updated version, links of the number of methods developed in the past few years have also been incorporated. This updated resource is built on responsive templates which is compatible with smartphones (mobile, iPhone, iPad, tablets etc.) and large screen gadgets. In summary, ccPDB 2.0 is a user-friendly web-based platform that provides comprehensive as well as updated information about datasets.

De novo protein structure prediction using ultra-fast molecular dynamics simulation

Modern genomics sequencing techniques have provided a massive amount of protein sequences, but experimental endeavor in determining protein structures is largely lagging far behind the vast and unexplored sequences. Apparently, computational biology is playing a more important role in protein structure prediction than ever. Here, we present a system of de novo predictor, termed NiDelta, building on a deep convolutional neural network and statistical potential enabling molecular dynamics simulation for modeling protein tertiary structure. Combining with evolutionary-based residue-contacts, the presented predictor can predict the tertiary structures of a number of target proteins with remarkable accuracy. The proposed approach is demonstrated by calculations on a set of eighteen large proteins from different fold classes. The results show that the ultra-fast molecular dynamics simulation could dramatically reduce the gap between the sequence and its structure at atom level, and it could also present high efficiency in protein structure determination if sparse experimental data is available.

RaptorX-Angle: real-value prediction of protein backbone dihedral angles through a hybrid method of clustering and deep learning

Background

Protein dihedral angles provide a detailed description of protein local conformation. Predicted dihedral angles can be used to narrow down the conformational space of the whole polypeptide chain significantly, thus aiding protein tertiary structure prediction. However, direct angle prediction from sequence alone is challenging.

Results

In this article, we present a novel method (named RaptorX-Angle) to predict real-valued angles by combining clustering and deep learning. Tested on a subset of PDB25 and the targets in the latest two Critical Assessment of protein Structure Prediction (CASP), our method outperforms the existing state-of-art method SPIDER2 in terms of Pearson Correlation Coefficient (PCC) and Mean Absolute Error (MAE). Our result also shows approximately linear relationship between the real prediction errors and our estimated bounds. That is, the real prediction error can be well approximated by our estimated bounds.

Conclusions

Our study provides an alternative and more accurate prediction of dihedral angles, which may facilitate protein structure prediction and functional study.

Electronic supplementary material

The online version of this article (10.1186/s12859-018-2065-x) contains supplementary material, which is available to authorized users.

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/.

Grid-based prediction of torsion angle probabilities of protein backbone and its application to discrimination of protein intrinsic disorder regions and selection of model structures

BACKGROUND

Protein structure can be described by backbone torsion angles: rotational angles about the N-Cα bond (φ) and the Cα-C bond (ψ) or the angle between Cαi-1-Cαi-Cαi + 1 (θ) and the rotational angle about the Cαi-Cαi + 1 bond (τ). Thus, their accurate prediction is useful for structure prediction and model refinement. Early methods predicted torsion angles in a few discrete bins whereas most recent methods have focused on prediction of angles in real, continuous values. Real value prediction, however, is unable to provide the information on probabilities of predicted angles.

RESULTS

Here, we propose to predict angles in fine grids of 5° by using deep learning neural networks. We found that this grid-based technique can yield 2-6% higher accuracy in predicting angles in the same 5° bin than existing prediction techniques compared. We further demonstrate the usefulness of predicted probabilities at given angle bins in discrimination of intrinsically disorder regions and in selection of protein models.

CONCLUSIONS

The proposed method may be useful for characterizing protein structure and disorder. The method is available at http://sparks-lab.org/server/SPIDER2/ as a part of SPIDER2 package.

Deep learning methods for protein torsion angle prediction

Background

Deep learning is one of the most powerful machine learning methods that has achieved the state-of-the-art performance in many domains. Since deep learning was introduced to the field of bioinformatics in 2012, it has achieved success in a number of areas such as protein residue-residue contact prediction, secondary structure prediction, and fold recognition. In this work, we developed deep learning methods to improve the prediction of torsion (dihedral) angles of proteins.

Results

We design four different deep learning architectures to predict protein torsion angles. The architectures including deep neural network (DNN) and deep restricted Boltzmann machine (DRBN), deep recurrent neural network (DRNN) and deep recurrent restricted Boltzmann machine (DReRBM) since the protein torsion angle prediction is a sequence related problem. In addition to existing protein features, two new features (predicted residue contact number and the error distribution of torsion angles extracted from sequence fragments) are used as input to each of the four deep learning architectures to predict phi and psi angles of protein backbone. The mean absolute error (MAE) of phi and psi angles predicted by DRNN, DReRBM, DRBM and DNN is about 20–21° and 29–30° on an independent dataset. The MAE of phi angle is comparable to the existing methods, but the MAE of psi angle is 29°, 2° lower than the existing methods. On the latest CASP12 targets, our methods also achieved the performance better than or comparable to a state-of-the art method.

Conclusions

Our experiment demonstrates that deep learning is a valuable method for predicting protein torsion angles. The deep recurrent network architecture performs slightly better than deep feed-forward architecture, and the predicted residue contact number and the error distribution of torsion angles extracted from sequence fragments are useful features for improving prediction accuracy.

Electronic supplementary material

The online version of this article (10.1186/s12859-017-1834-2) contains supplementary material, which is available to authorized users.

Identification of metal ion binding sites based on amino acid sequences

The identification of metal ion binding sites is important for protein function annotation and the design of new drug molecules. This study presents an effective method of analyzing and identifying the binding residues of metal ions based solely on sequence information. Ten metal ions were extracted from the BioLip database: Zn2+, Cu2+, Fe2+, Fe3+, Ca2+, Mg2+, Mn2+, Na+, K+ and Co2+. The analysis showed that Zn2+, Cu2+, Fe2+, Fe3+, and Co2+ were sensitive to the conservation of amino acids at binding sites, and promising results can be achieved using the Position Weight Scoring Matrix algorithm, with an accuracy of over 79.9% and a Matthews correlation coefficient of over 0.6. The binding sites of other metals can also be accurately identified using the Support Vector Machine algorithm with multifeature parameters as input. In addition, we found that Ca2+ was insensitive to hydrophobicity and hydrophilicity information and Mn2+ was insensitive to polarization charge information. An online server was constructed based on the framework of the proposed method and is freely available at http://60.31.198.140:8081/metal/HomePage/HomePage.html.

Analysis of the substrate specificity of Factor VII activating protease (FSAP) and design of specific and sensitive peptide substrates

Factor VII (FVII) activating protease (FSAP) is a circulating serine protease that is likely to be involved in a number of disease conditions such as stroke, atherosclerosis, liver fibrosis, thrombosis and cancer. To date, no systematic information is available about the substrate specificity of FSAP. Applying phage display and positional scanning substrate combinatorial library (PS-SCL) approaches we have characterised the specificity of FSAP towards small peptides. Results were evaluated in the context of known protein substrates as well as molecular modelling of the peptides in the active site of FSAP. The representative FSAP-cleaved sequence obtained from the phage display method was Val-Leu-Lys-Arg-Ser (P4-P1'). The sequence X-Lys/Arg-Nle-Lys/Arg (P4-P1) was derived from the PS-SCL method. These results show a predilection for cleavage at a cluster of basic amino acids on the nonprime side. Quenched fluorescent substrate (Ala-Lys-Nle-Arg-AMC) (amino methyl coumarin) and (Ala-Leu-Lys-Arg-AMC) had a higher selectivity for FSAP compared to other proteases from the hemostasis system. These substrates could be used to measure FSAP activity in a complex biological system such as plasma. In histone-treated plasma there was a specific activation of pro-FSAP as validated by the use of an FSAP inhibitory antibody, corn trypsin inhibitor to inhibit Factor XIIa and hirudin to inhibit thrombin, which may account for some of the haemostasis-related effects of histones. These results will aid the development of further selective FSAP activity probes as well as specific inhibitors that will help to increase the understanding of the functions of FSAP in vivo.

SPIDER2: A Package to Predict Secondary Structure, Accessible Surface Area, and Main-Chain Torsional Angles by Deep Neural Networks

Predicting one-dimensional structure properties has played an important role to improve prediction of protein three-dimensional structures and functions. The most commonly predicted properties are secondary structure and accessible surface area (ASA) representing local and nonlocal structural characteristics, respectively. Secondary structure prediction is further complemented by prediction of continuous main-chain torsional angles. Here we describe a newly developed method SPIDER2 that utilizes three iterations of deep learning neural networks to improve the prediction accuracy of several structural properties simultaneously. For an independent test set of 1199 proteins SPIDER2 achieves 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. The method provides state-of-the-art, all-in-one accurate prediction of local structure and solvent accessible surface area. The method is implemented, as a webserver along with a standalone package that are available in our website: http://sparks-lab.org .

An Evolution-Based Approach to De Novo Protein Design

EvoDesign is a computational algorithm that allows the rapid creation of new protein sequences that are compatible with specific protein structures. As such, it can be used to optimize protein stability, to resculpt the protein surface to eliminate undesired protein-protein interactions, and to optimize protein-protein binding. A major distinguishing feature of EvoDesign in comparison to other protein design programs is the use of evolutionary information in the design process to guide the sequence search toward native-like sequences known to adopt structurally similar folds as the target. The observed frequencies of amino acids in specific positions in the structure in the form of structural profiles collected from proteins with similar folds and complexes with similar interfaces can implicitly capture many subtle effects that are essential for correct folding and protein-binding interactions. As a result of the inclusion of evolutionary information, the sequences designed by EvoDesign have native-like folding and binding properties not seen by other physics-based design methods. In this chapter, we describe how EvoDesign can be used to redesign proteins with a focus on the computational and experimental procedures that can be used to validate the designs.

Backbone Dihedral Angle Prediction

More than two decades of research have enabled dihedral angle predictions at an accuracy that makes them an interesting alternative or supplement to secondary structure prediction that provides detailed local structure information for every residue of a protein. The evolution of dihedral angle prediction methods is closely linked to advancements in machine learning and other relevant technologies. Consequently recent improvements in large-scale training of deep neural networks have led to the best method currently available, which achieves a mean absolute error of 19° for phi, and 30° for psi. This performance opens interesting perspectives for the application of dihedral angle prediction in the comparison, prediction, and design of protein structures.

Protein ligand-specific binding residue predictions by an ensemble classifier

BACKGROUND

Prediction of ligand binding sites is important to elucidate protein functions and is helpful for drug design. Although much progress has been made, many challenges still need to be addressed. Prediction methods need to be carefully developed to account for chemical and structural differences between ligands.

RESULTS

In this study, we present ligand-specific methods to predict the binding sites of protein-ligand interactions. First, a sequence-based method is proposed that only extracts features from protein sequence information, including evolutionary conservation scores and predicted structure properties. An improved AdaBoost algorithm is applied to address the serious imbalance problem between the binding and non-binding residues. Then, a combined method is proposed that combines the current template-free method and four other well-established template-based methods. The above two methods predict the ligand binding sites along the sequences using a ligand-specific strategy that contains metal ions, acid radical ions, nucleotides and ferroheme. Testing on a well-established dataset showed that the proposed sequence-based method outperformed the profile-based method by 4-19% in terms of the Matthews correlation coefficient on different ligands. The combined method outperformed each of the individual methods, with an improvement in the average Matthews correlation coefficients of 5.55% over all ligands. The results also show that the ligand-specific methods significantly outperform the general-purpose methods, which confirms the necessity of developing elaborate ligand-specific methods for ligand binding site prediction.

CONCLUSIONS

Two efficient ligand-specific binding site predictors are presented. The standalone package is freely available for academic usage at http://dase.ecnu.edu.cn/qwdong/TargetCom/TargetCom_standalone.tar.gz  or request upon the corresponding author.

Recognizing metal and acid radical ion-binding sites by integrating ab initio modeling with template-based transferals

MOTIVATION

More than half of proteins require binding of metal and acid radical ions for their structure and function. Identification of the ion-binding locations is important for understanding the biological functions of proteins. Due to the small size and high versatility of the metal and acid radical ions, however, computational prediction of their binding sites remains difficult.

RESULTS

We proposed a new ligand-specific approach devoted to the binding site prediction of 13 metal ions (Zn²⁺, Cu²⁺, Fe²⁺, Fe³⁺, Ca²⁺, Mg²⁺, Mn²⁺, Na⁺, K⁺) and acid radical ion ligands (CO3^2-, NO2^-, SO4^2-, PO4^3-) that are most frequently seen in protein databases. A sequence-based ab initio model is first trained on sequence profiles, where a modified AdaBoost algorithm is extended to balance binding and non-binding residue samples. A composite method IonCom is then developed to combine the ab initio model with multiple threading alignments for further improving the robustness of the binding site predictions. The pipeline was tested using 5-fold cross validations on a comprehensive set of 2,100 non-redundant proteins bound with 3,075 small ion ligands. Significant advantage was demonstrated compared with the state of the art ligand-binding methods including COACH and TargetS for high-accuracy ion-binding site identification. Detailed data analyses show that the major advantage of IonCom lies at the integration of complementary ab initio and template-based components. Ion-specific feature design and binding library selection also contribute to the improvement of small ion ligand binding predictions.

AVAILABILITY AND IMPLEMENTATION

http://zhanglab.ccmb.med.umich.edu/IonCom CONTACT: hxz@imut.edu.cn or zhng@umich.eduSupplementary information: Supplementary data are available at Bioinformatics online.

STRUM: structure-based prediction of protein stability changes upon single-point mutation

MOTIVATION

Mutations in human genome are mainly through single nucleotide polymorphism, some of which can affect stability and function of proteins, causing human diseases. Several methods have been proposed to predict the effect of mutations on protein stability; but most require features from experimental structure. Given the fast progress in protein structure prediction, this work explores the possibility to improve the mutation-induced stability change prediction using low-resolution structure modeling.

RESULTS

We developed a new method (STRUM) for predicting stability change caused by single-point mutations. Starting from wild-type sequences, 3D models are constructed by the iterative threading assembly refinement (I-TASSER) simulations, where physics- and knowledge-based energy functions are derived on the I-TASSER models and used to train STRUM models through gradient boosting regression. STRUM was assessed by 5-fold cross validation on 3421 experimentally determined mutations from 150 proteins. The Pearson correlation coefficient (PCC) between predicted and measured changes of Gibbs free-energy gap, ΔΔG, upon mutation reaches 0.79 with a root-mean-square error 1.2 kcal/mol in the mutation-based cross-validations. The PCC reduces if separating training and test mutations from non-homologous proteins, which reflects inherent correlations in the current mutation sample. Nevertheless, the results significantly outperform other state-of-the-art methods, including those built on experimental protein structures. Detailed analyses show that the most sensitive features in STRUM are the physics-based energy terms on I-TASSER models and the conservation scores from multiple-threading template alignments. However, the ΔΔG prediction accuracy has only a marginal dependence on the accuracy of protein structure models as long as the global fold is correct. These data demonstrate the feasibility to use low-resolution structure modeling for high-accuracy stability change prediction upon point mutations.

AVAILABILITY AND IMPLEMENTATION

http://zhanglab.ccmb.med.umich.edu/STRUM/ CONTACT: qiang@suda.edu.cn and zhng@umich.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A neural network learning approach for improving the prediction of residue depth based on sequence-derived features

Residue depth is a solvent exposure measure that quantitatively describes the depth of a residue from the protein surface.

Adaptive firefly algorithm: parameter analysis and its application

As a nature-inspired search algorithm, firefly algorithm (FA) has several control parameters, which may have great effects on its performance. In this study, we investigate the parameter selection and adaptation strategies in a modified firefly algorithm — adaptive firefly algorithm (AdaFa). There are three strategies in AdaFa including (1) a distance-based light absorption coefficient; (2) a gray coefficient enhancing fireflies to share difference information from attractive ones efficiently; and (3) five different dynamic strategies for the randomization parameter. Promising selections of parameters in the strategies are analyzed to guarantee the efficient performance of AdaFa. AdaFa is validated over widely used benchmark functions, and the numerical experiments and statistical tests yield useful conclusions on the strategies and the parameter selections affecting the performance of AdaFa. When applied to the real-world problem — protein tertiary structure prediction, the results demonstrated improved variants can rebuild the tertiary structure with the average root mean square deviation less than 0.4Å and 1.5Å from the native constrains with noise free and 10% Gaussian white noise.

Evaluation of protein dihedral angle prediction methods

Tertiary structure prediction of a protein from its amino acid sequence is one of the major challenges in the field of bioinformatics. Hierarchical approach is one of the persuasive techniques used for predicting protein tertiary structure, especially in the absence of homologous protein structures. In hierarchical approach, intermediate states are predicted like secondary structure, dihedral angles, C^α-C^α distance bounds, etc. These intermediate states are used to restraint the protein backbone and assist its correct folding. In the recent years, several methods have been developed for predicting dihedral angles of a protein, but it is difficult to conclude which method is better than others. In this study, we benchmarked the performance of dihedral prediction methods ANGLOR and SPINE X on various datasets, including independent datasets. TANGLE dihedral prediction method was not benchmarked (due to unavailability of its standalone) and was compared with SPINE X and ANGLOR on only ANGLOR dataset on which TANGLE has reported its results. It was observed that SPINE X performed better than ANGLOR and TANGLE, especially in case of prediction of dihedral angles of glycine and proline residues. The analysis suggested that angle shifting was the foremost reason of better performance of SPINE X. We further evaluated the performance of the methods on independent ccPDB30 dataset and observed that SPINE X performed better than ANGLOR.

A comparative assessment and analysis of 20 representative sequence alignment methods for protein structure prediction

Protein sequence alignment is essential for template-based protein structure prediction and function annotation. We collect 20 sequence alignment algorithms, 10 published and 10 newly developed, which cover all representative sequence- and profile-based alignment approaches. These algorithms are benchmarked on 538 non-redundant proteins for protein fold-recognition on a uniform template library. Results demonstrate dominant advantage of profile-profile based methods, which generate models with average TM-score 26.5% higher than sequence-profile methods and 49.8% higher than sequence-sequence alignment methods. There is no obvious difference in results between methods with profiles generated from PSI-BLAST PSSM matrix and hidden Markov models. Accuracy of profile-profile alignments can be further improved by 9.6% or 21.4% when predicted or native structure features are incorporated. Nevertheless, TM-scores from profile-profile methods including experimental structural features are still 37.1% lower than that from TM-align, demonstrating that the fold-recognition problem cannot be solved solely by improving accuracy of structure feature predictions.

From local structure to a global framework: recognition of protein folds

Protein folding has been a major area of research for many years. Nonetheless, the mechanisms leading to the formation of an active biological fold are still not fully apprehended. The huge amount of available sequence and structural information provides hints to identify the putative fold for a given sequence. Indeed, protein structures prefer a limited number of local backbone conformations, some being characterized by preferences for certain amino acids. These preferences largely depend on the local structural environment. The prediction of local backbone conformations has become an important factor to correctly identifying the global protein fold. Here, we review the developments in the field of local structure prediction and especially their implication in protein fold recognition.

Confidence-guided local structure prediction with HHfrag

We present a method to assess the reliability of local structure prediction from sequence. We introduce a greedy algorithm for filtering and enrichment of dynamic fragment libraries, compiled with remote-homology detection methods such as HHfrag. After filtering false hits at each target position, we reduce the fragment library to a minimal set of representative fragments, which are guaranteed to have correct local structure in regions of detectable conservation. We demonstrate that the location of conserved motifs in a protein sequence can be predicted by examining the recurrence and structural homogeneity of detected fragments. The resulting confidence score correlates with the local RMSD of the representative fragments and allows us to predict torsion angles from sequence with better accuracy compared to existing machine learning methods.

EvoDesign: De novo protein design based on structural and evolutionary profiles

Protein design aims to identify new protein sequences of desirable structure and biological function. Most current de novo protein design methods rely on physics-based force fields to search for low free-energy states following Anfinsen’s thermodynamic hypothesis. A major obstacle of such approaches is the inaccuracy of the force field design, which cannot accurately describe the atomic interactions or distinguish correct folds. We developed a new web server, EvoDesign, to design optimal protein sequences of given scaffolds along with multiple sequence and structure-based features to assess the foldability and goodness of the designs. EvoDesign uses an evolution-profile–based Monte Carlo search with the profiles constructed from homologous structure families in the Protein Data Bank. A set of local structure features, including secondary structure, torsion angle and solvation, are predicted by single-sequence neural-network training and used to smooth the sequence motif and accommodate the physicochemical packing. The EvoDesign algorithm has been extensively tested in large-scale protein design experiments, which demonstrate enhanced foldability and structural stability of designed sequences compared with the physics-based designing methods. The EvoDesign server is freely available at http://zhanglab.ccmb.med.umich.edu/EvoDesign.

Toward optimal fragment generations for ab initio protein structure assembly

Fragment assembly using structural motifs excised from other solved proteins has shown to be an efficient method for ab initio protein-structure prediction. However, how to construct accurate fragments, how to derive optimal restraints from fragments, and what the best fragment length is are the basic issues yet to be systematically examined. In this work, we developed a gapless-threading method to generate position-specific structure fragments. Distance profiles and torsion angle pairs are then derived from the fragments by statistical consistency analysis, which achieved comparable accuracy with the machine-learning-based methods although the fragments were taken from unrelated proteins. When measured by both accuracies of the derived distance profiles and torsion angle pairs, we come to a consistent conclusion that the optimal fragment length for structural assembly is around 10, and at least 100 fragments at each location are needed to achieve optimal structure assembly. The distant profiles and torsion angle pairs as derived by the fragments have been successfully used in QUARK for ab initio protein structure assembly and are provided by the QUARK online server at http://zhanglab.ccmb. med.umich.edu/QUARK/.