Predicting residue-residue contacts using random forest models

Navigating the landscape of enzyme design: from molecular simulations to machine learning

Global environmental issues and sustainable development call for new technologies for fine chemical synthesis and waste valorization. Biocatalysis has attracted great attention as the alternative to the traditional organic synthesis. However, it is challenging to navigate the vast sequence space to identify those proteins with admirable biocatalytic functions. The recent development of deep-learning based structure prediction methods such as AlphaFold2 reinforced by different computational simulations or multiscale calculations has largely expanded the 3D structure databases and enabled structure-based design. While structure-based approaches shed light on site-specific enzyme engineering, they are not suitable for large-scale screening of potential biocatalysts. Effective utilization of big data using machine learning techniques opens up a new era for accelerated predictions. Here, we review the approaches and applications of structure-based and machine-learning guided enzyme design. We also provide our view on the challenges and perspectives on effectively employing enzyme design approaches integrating traditional molecular simulations and machine learning, and the importance of database construction and algorithm development in attaining predictive ML models to explore the sequence fitness landscape for the design of admirable biocatalysts.

Vaccine design and development: Exploring the interface with computational biology and AI

Computational biology involves applying computer science and informatics techniques in biology to understand complex biological data. It allows us to collect, connect, and analyze biological data at a large scale and build predictive models. In the twenty first century, computational resources along with Artificial Intelligence (AI) have been widely used in various fields of biological sciences such as biochemistry, structural biology, immunology, microbiology, and genomics to handle massive data for decision-making, including in applications such as drug design and vaccine development, one of the major areas of focus for human and animal welfare. The knowledge of available computational resources and AI-enabled tools in vaccine design and development can improve our ability to conduct cutting-edge research. Therefore, this review article aims to summarize important computational resources and AI-based tools. Further, the article discusses the various applications and limitations of AI tools in vaccine development.

Proximal aortic repair in dialysis patients: A national database analysis

OBJECTIVES

Dialysis is a well-established risk factor for morbidity and mortality after cardiovascular procedures. However, little is known regarding the outcomes of proximal aortic surgery in this high-risk cohort.

METHODS

Perioperative (in-hospital or 30-day mortality) and 10-year outcomes were analyzed for all the patients who underwent open proximal aortic repair with the diagnosis of nonruptured thoracic aortic aneurysm (aneurysm, n = 325) or type A aortic dissection (dissection, n = 461) from 1987 to 2015 using the US Renal Data System database.

RESULTS

In patients with aneurysm, perioperative mortality was 12.6%. The 10-year mortality was 81% ± 3%. Age 65 years or more (hazard ratio [HR], 1.35; 95% confidence interval [CI], 1.03 to 1.78; P = .03), chronic obstructive pulmonary disease (HR, 1.68; 95% CI, 1.01-2.82; P = .047), and Black race (HR, 1.46; 95% CI, 1.09-1.97; P = .01) were independently associated with worse 10-year mortality. In patients with dissection, perioperative mortality was 24.3% and 10-year mortality was 87.9% ± 2.2%. Age 65 years or more (HR, 1.49; 95% CI, 1.19-1.86; P < .001), congestive heart failure (HR, 1.39; 95% CI, 1.11-2.57; P = .004), and diabetes mellitus as the cause of dialysis (HR, 1.75; 95% CI, 1.2-2.57; P = .004) were independently associated with worse 10-year mortality. Black race (HR, 0.74; 95% CI, 0.6-0.92; P = .008) was associated with a better outcome.

CONCLUSIONS

We described challenging perioperative and 10-year outcomes for dialysis patients undergoing proximal aortic repair. The present study suggests the need for careful patient selection in the elective repair of proximal aortic aneurysm for dialysis-dependent patients, whereas it affirms the feasibility of emergency surgery for acute type A aortic dissections.

New Labeling Methods for Deep Learning Real-Valued Inter-Residue Distance Prediction

BACKGROUND

Much of the recent success in protein structure prediction has been a result of accurate protein contact prediction-a binary classification problem. Dozens of methods, built from various types of machine learning and deep learning algorithms, have been published over the last two decades for predicting contacts. Recently, many groups, including Google DeepMind, have demonstrated that reformulating the problem as a multi-class classification problem is a more promising direction to pursue. As an alternative approach, we recently proposed real-valued distance predictions, formulating the problem as a regression problem. The nuances of protein 3D structures make this formulation appropriate, allowing predictions to reflect inter-residue distances in nature. Despite these promises, the accurate prediction of real-valued distances remains relatively unexplored; possibly due to classification being better suited to machine and deep learning algorithms.

METHODS

Can regression methods be designed to predict real-valued distances as precise as binary contacts? To investigate this, we propose multiple novel methods of input label engineering, which is different from feature engineering, with the goal of optimizing the distribution of distances to cater to the loss function of the deep-learning model. Since an important utility of predicted contacts or distances is to build three-dimensional models, we also tested if predicted distances can reconstruct more accurate models than contacts.

RESULTS

Our results demonstrate, for the first time, that deep learning methods for real-valued protein distance prediction can deliver distances as precise as binary classification methods. When using an optimal distance transformation function on the standard PSICOV dataset consisting of 150 representative proteins, the precision of 'top-all' long-range contacts improves from 60.9% to 61.4% when predicting real-valued distances instead of contacts. When building three-dimensional models we observed an average TM-score increase from 0.61 to 0.72, highlighting the advantage of predicting real-valued distances.

Deep learning methods for 3D structural proteome and interactome modeling

Bolstered by recent methodological and hardware advances, deep learning has increasingly been applied to biological problems and structural proteomics. Such approaches have achieved remarkable improvements over traditional machine learning methods in tasks ranging from protein contact map prediction to protein folding, prediction of protein-protein interaction interfaces, and characterization of protein-drug binding pockets. In particular, emergence of ab initio protein structure prediction methods including AlphaFold2 has revolutionized protein structural modeling. From a protein function perspective, numerous deep learning methods have facilitated deconvolution of the exact amino acid residues and protein surface regions responsible for binding other proteins or small molecule drugs. In this review, we provide a comprehensive overview of recent deep learning methods applied in structural proteomics.

Highly Similar Sequence and Structure Yet Different Biophysical Behavior: A Computational Study of Two Triosephosphate Isomerases

Homodimeric triosephosphate isomerases (TIMs) from Trypanosoma cruzi (TcTIM) and Trypanosoma brucei (TbTIM) have markedly similar amino-acid sequences and three-dimensional structures. However, several of their biophysical parameters, such as their susceptibility to sulfhydryl agents and their reactivation speed after being denatured, have significant differences. The causes of these differences were explored with microsecond-scale molecular dynamics (MD) simulations of three different TIM proteins: TcTIM, TbTIM, and a chimeric protein, Mut1. We examined their electrostatic interactions and explored the impact of simulation length on them. The same salt bridge between catalytic residues Lys 14 and Glu 98 was observed in all three proteins, but key differences were found in other interactions that the catalytic amino acids form. In particular, a cation-π interaction between catalytic amino acids Lys 14 and His 96 and both a salt bridge and a hydrogen bond between catalytic Glu 168 and residue Arg 100 were only observed in TcTIM. Furthermore, although TcTIM forms less hydrogen bonds than TbTIM and Mut1, its hydrogen bond network spans almost the entire protein, connecting the residues in both monomers. This work provides new insight into the mechanisms that give rise to the different behavior of these proteins. The results also show the importance of long simulations.

DeepHelicon: Accurate prediction of inter-helical residue contacts in transmembrane proteins by residual neural networks

Accurate prediction of amino acid residue contacts is an important prerequisite for generating high-quality 3D models of transmembrane (TM) proteins. While a large number of compositional, evolutionary, and structural properties of proteins can be used to train contact prediction methods, recent research suggests that coevolution between residues provides the strongest indication of their spatial proximity. We have developed a deep learning approach, DeepHelicon, to predict inter-helical residue contacts in TM proteins by considering only coevolutionary features. DeepHelicon comprises a two-stage supervised learning process by residual neural networks for a gradual refinement of contact maps, followed by variance reduction by an ensemble of models. We present a benchmark study of 12 contact predictors and conclude that DeepHelicon together with the two other state-of-the-art methods DeepMetaPSICOV and Membrain2 outperforms the 10 remaining algorithms on all datasets and at all settings. On a set of 44 TM proteins with an average length of 388 residues DeepHelicon achieves the best performance among all benchmarked methods in predicting the top L/5 and L/2 inter-helical contacts, with the mean precision of 87.42% and 77.84%, respectively. On a set of 57 relatively small TM proteins with an average length of 298 residues DeepHelicon ranks second best after DeepMetaPSICOV. DeepHelicon produces the most accurate predictions for large proteins with more than 10 transmembrane helices. Coevolutionary features alone allow to predict inter-helical residue contacts with an accuracy sufficient for generating acceptable 3D models for up to 30% of proteins using a fully automated modeling method such as CONFOLD2.

Predicting protein residue-residue contacts using random forests and deep networks

BACKGROUND

The ability to predict which pairs of amino acid residues in a protein are in contact with each other offers many advantages for various areas of research that focus on proteins. For example, contact prediction can be used to reduce the computational complexity of predicting the structure of proteins and even to help identify functionally important regions of proteins. These predictions are becoming especially important given the relatively low number of experimentally determined protein structures compared to the amount of available protein sequence data.

RESULTS

Here we have developed and benchmarked a set of machine learning methods for performing residue-residue contact prediction, including random forests, direct-coupling analysis, support vector machines, and deep networks (stacked denoising autoencoders). These methods are able to predict contacting residue pairs given only the amino acid sequence of a protein. According to our own evaluations performed at a resolution of +/- two residues, the predictors we trained with the random forest algorithm were our top performing methods with average top 10 prediction accuracy scores of 85.13% (short range), 74.49% (medium range), and 54.49% (long range). Our ensemble models (stacked denoising autoencoders combined with support vector machines) were our best performing deep network predictors and achieved top 10 prediction accuracy scores of 75.51% (short range), 60.26% (medium range), and 43.85% (long range) using the same evaluation. These tests were blindly performed on targets from the CASP11 dataset; and the results suggested that our models achieved comparable performance to contact predictors developed by groups that participated in CASP11.

CONCLUSIONS

Due to the challenging nature of contact prediction, it is beneficial to develop and benchmark a variety of different prediction methods. Our work has produced useful tools with a simple interface that can provide contact predictions to users without requiring a lengthy installation process. In addition to this, we have released our C++ implementation of the direct-coupling analysis method as a standalone software package. Both this tool and our RFcon web server are freely available to the public at http://dna.cs.miami.edu/RFcon /.

Protein Inter-Residue Contacts Prediction: Methods, Performances and Applications

Background:

Protein inter-residue contacts prediction play an important role in the field of protein structure and function research. As a low-dimensional representation of protein tertiary structure, protein inter-residue contacts could greatly help de novo protein structure prediction methods to reduce the conformational search space. Over the past two decades, various methods have been developed for protein inter-residue contacts prediction.

Objective:

We provide a comprehensive and systematic review of protein inter-residue contacts prediction methods.

Results:

Protein inter-residue contacts prediction methods are roughly classified into five categories: correlated mutations methods, machine-learning methods, fusion methods, templatebased methods and 3D model-based methods. In this paper, firstly we describe the common definition of protein inter-residue contacts and show the typical application of protein inter-residue contacts. Then, we present a comprehensive review of the three main categories for protein interresidue contacts prediction: correlated mutations methods, machine-learning methods and fusion methods. Besides, we analyze the constraints for each category. Furthermore, we compare several representative methods on the CASP11 dataset and discuss performances of these methods in detail.

Conclusion:

Correlated mutations methods achieve better performances for long-range contacts, while the machine-learning method performs well for short-range contacts. Fusion methods could take advantage of the machine-learning and correlated mutations methods. Employing more effective fusion strategy could be helpful to further improve the performances of fusion methods.

A large-scale comparative assessment of methods for residue-residue contact prediction

Sequence-based prediction of residue-residue contact in proteins becomes increasingly more important for improving protein structure prediction in the big data era. In this study, we performed a large-scale comparative assessment of 15 locally installed contact predictors. To assess these methods, we collected a big data set consisting of 680 nonredundant proteins covering different structural classes and target difficulties. We investigated a wide range of factors that may influence the precision of contact prediction, including target difficulty, structural class, the alignment depth and distribution of contact pairs in a protein structure. We found that: (1) the machine learning-based methods outperform the direct-coupling-based methods for short-range contact prediction, while the latter are significantly better for long-range contact prediction. The consensus-based methods, which combine machine learning and direct-coupling methods, perform the best. (2) The target difficulty does not have clear influence on the machine learning-based methods, while it does affect the direct-coupling and consensus-based methods significantly. (3) The alignment depth has relatively weak effect on the machine learning-based methods. However, for the direct-coupling-based methods and consensus-based methods, the predicted contacts for targets with deeper alignment tend to be more accurate. (4) All methods perform relatively better on β and α + β proteins than on α proteins. (5) Residues buried in the core of protein structure are more prone to be in contact than residues on the surface (22 versus 6%). We believe these are useful results for guiding future development of new approach to contact prediction.

Unified Deep Learning Architecture for Modeling Biology Sequence

Prediction of the spatial structure or function of biological macromolecules based on their sequences remains an important challenge in bioinformatics. When modeling biological sequences using traditional sequencing models, long-range interaction, complicated and variable output of labeled structures, and variable length of biological sequences usually lead to different solutions on a case-by-case basis. This study proposed a unified deep learning architecture based on long short-term memory or a gated recurrent unit to capture long-range interactions. The architecture designs the optional reshape operator to adapt to the diversity of the output labels and implements a training algorithm to support the training of sequence models capable of processing variable-length sequences. The merging and pooling operators enhances the ability of capturing short-range interactions between basic units of biological sequences. The proposed deep-learning architecture and its training algorithm might be capable of solving currently variable biological sequence-modeling problems under a unified framework. We validated the model on one of the most difficult biological sequence-modeling problems, protein residue interaction prediction. The results indicate that the accuracy of obtaining the residue interactions of the model exceeded popular approaches by 10 percent on multiple widely-used benchmarks.

Finding the needle in the haystack: towards solving the protein-folding problem computationally

Prediction of protein tertiary structures from amino acid sequence and understanding the mechanisms of how proteins fold, collectively known as "the protein folding problem," has been a grand challenge in molecular biology for over half a century. Theories have been developed that provide us with an unprecedented understanding of protein folding mechanisms. However, computational simulation of protein folding is still difficult, and prediction of protein tertiary structure from amino acid sequence is an unsolved problem. Progress toward a satisfying solution has been slow due to challenges in sampling the vast conformational space and deriving sufficiently accurate energy functions. Nevertheless, several techniques and algorithms have been adopted to overcome these challenges, and the last two decades have seen exciting advances in enhanced sampling algorithms, computational power and tertiary structure prediction methodologies. This review aims at summarizing these computational techniques, specifically conformational sampling algorithms and energy approximations that have been frequently used to study protein-folding mechanisms or to de novo predict protein tertiary structures. We hope that this review can serve as an overview on how the protein-folding problem can be studied computationally and, in cases where experimental approaches are prohibitive, help the researcher choose the most relevant computational approach for the problem at hand. We conclude with a summary of current challenges faced and an outlook on potential future directions.

Forecasting residue-residue contact prediction accuracy

Motivation

Apart from meta-predictors, most of today's methods for residue-residue contact prediction are based entirely on Direct Coupling Analysis (DCA) of correlated mutations in multiple sequence alignments (MSAs). These methods are on average ∼40% correct for the 100 strongest predicted contacts in each protein. The end-user who works on a single protein of interest will not know if predictions are either much more or much less correct than 40%, which is especially a problem if contacts are predicted to steer experimental research on that protein.

Results

We designed a regression model that forecasts the accuracy of residue-residue contact prediction for individual proteins with an average error of 7 percentage points. Contacts were predicted with two DCA methods (gplmDCA and PSICOV). The models were built on parameters that describe the MSA, the predicted secondary structure, the predicted solvent accessibility and the contact prediction scores for the target protein. Results show that our models can be also applied to the meta-methods, which was tested on RaptorX.

Availability and implementation

All data and scripts are available from http://comprec-lin.iiar.pwr.edu.pl/dcaQ/.

Contact

malgorzata.kotulska@pwr.edu.pl.

Supplementary information

Supplementary data are available at Bioinformatics online.

RRCRank: a fusion method using rank strategy for residue-residue contact prediction

Background

In structural biology area, protein residue-residue contacts play a crucial role in protein structure prediction. Some researchers have found that the predicted residue-residue contacts could effectively constrain the conformational search space, which is significant for de novo protein structure prediction. In the last few decades, related researchers have developed various methods to predict residue-residue contacts, especially, significant performance has been achieved by using fusion methods in recent years. In this work, a novel fusion method based on rank strategy has been proposed to predict contacts. Unlike the traditional regression or classification strategies, the contact prediction task is regarded as a ranking task. First, two kinds of features are extracted from correlated mutations methods and ensemble machine-learning classifiers, and then the proposed method uses the learning-to-rank algorithm to predict contact probability of each residue pair.

Results

First, we perform two benchmark tests for the proposed fusion method (RRCRank) on CASP11 dataset and CASP12 dataset respectively. The test results show that the RRCRank method outperforms other well-developed methods, especially for medium and short range contacts. Second, in order to verify the superiority of ranking strategy, we predict contacts by using the traditional regression and classification strategies based on the same features as ranking strategy. Compared with these two traditional strategies, the proposed ranking strategy shows better performance for three contact types, in particular for long range contacts. Third, the proposed RRCRank has been compared with several state-of-the-art methods in CASP11 and CASP12. The results show that the RRCRank could achieve comparable prediction precisions and is better than three methods in most assessment metrics.

Conclusions

The learning-to-rank algorithm is introduced to develop a novel rank-based method for the residue-residue contact prediction of proteins, which achieves state-of-the-art performance based on the extensive assessment.

Electronic supplementary material

The online version of this article (10.1186/s12859-017-1811-9) contains supplementary material, which is available to authorized users.

Elastic network model of learned maintained contacts to predict protein motion

We present a novel elastic network model, lmcENM, to determine protein motion even for localized functional motions that involve substantial changes in the protein's contact topology. Existing elastic network models assume that the contact topology remains unchanged throughout the motion and are thus most appropriate to simulate highly collective function-related movements. lmcENM uses machine learning to differentiate breaking from maintained contacts. We show that lmcENM accurately captures functional transitions unexplained by the classical ENM and three reference ENM variants, while preserving the simplicity of classical ENM. We demonstrate the effectiveness of our approach on a large set of proteins covering different motion types. Our results suggest that accurately predicting a "deformation-invariant" contact topology offers a promising route to increase the general applicability of ENMs. We also find that to correctly predict this contact topology a combination of several features seems to be relevant which may vary slightly depending on the protein. Additionally, we present case studies of two biologically interesting systems, Ferric Citrate membrane transporter FecA and Arachidonate 15-Lipoxygenase.

EPSILON-CP: using deep learning to combine information from multiple sources for protein contact prediction

BACKGROUND

Accurately predicted contacts allow to compute the 3D structure of a protein. Since the solution space of native residue-residue contact pairs is very large, it is necessary to leverage information to identify relevant regions of the solution space, i.e. correct contacts. Every additional source of information can contribute to narrowing down candidate regions. Therefore, recent methods combined evolutionary and sequence-based information as well as evolutionary and physicochemical information. We develop a new contact predictor (EPSILON-CP) that goes beyond current methods by combining evolutionary, physicochemical, and sequence-based information. The problems resulting from the increased dimensionality and complexity of the learning problem are combated with a careful feature analysis, which results in a drastically reduced feature set. The different information sources are combined using deep neural networks.

RESULTS

On 21 hard CASP11 FM targets, EPSILON-CP achieves a mean precision of 35.7% for top- L/10 predicted long-range contacts, which is 11% better than the CASP11 winning version of MetaPSICOV. The improvement on 1.5L is 17%. Furthermore, in this study we find that the amino acid composition, a commonly used feature, is rendered ineffective in the context of meta approaches. The size of the refined feature set decreased by 75%, enabling a significant increase in training data for machine learning, contributing significantly to the observed improvements.

CONCLUSIONS

Exploiting as much and diverse information as possible is key to accurate contact prediction. Simply merging the information introduces new challenges. Our study suggests that critical feature analysis can improve the performance of contact prediction methods that combine multiple information sources. EPSILON-CP is available as a webservice: http://compbio.robotics.tu-berlin.de/epsilon/.

Deep learning for computational chemistry

The rise and fall of artificial neural networks is well documented in the scientific literature of both computer science and computational chemistry. Yet almost two decades later, we are now seeing a resurgence of interest in deep learning, a machine learning algorithm based on multilayer neural networks. Within the last few years, we have seen the transformative impact of deep learning in many domains, particularly in speech recognition and computer vision, to the extent that the majority of expert practitioners in those field are now regularly eschewing prior established models in favor of deep learning models. In this review, we provide an introductory overview into the theory of deep neural networks and their unique properties that distinguish them from traditional machine learning algorithms used in cheminformatics. By providing an overview of the variety of emerging applications of deep neural networks, we highlight its ubiquity and broad applicability to a wide range of challenges in the field, including quantitative structure activity relationship, virtual screening, protein structure prediction, quantum chemistry, materials design, and property prediction. In reviewing the performance of deep neural networks, we observed a consistent outperformance against non-neural networks state-of-the-art models across disparate research topics, and deep neural network-based models often exceeded the "glass ceiling" expectations of their respective tasks. Coupled with the maturity of GPU-accelerated computing for training deep neural networks and the exponential growth of chemical data on which to train these networks on, we anticipate that deep learning algorithms will be a valuable tool for computational chemistry. © 2017 Wiley Periodicals, Inc.

In silico identification of enhancers on the basis of a combination of transcription factor binding motif occurrences

Enhancers interact with gene promoters and form chromatin looping structures that serve important functions in various biological processes, such as the regulation of gene transcription and cell differentiation. However, enhancers are difficult to identify because they generally do not have fixed positions or consensus sequence features, and biological experiments for enhancer identification are costly in terms of labor and expense. In this work, several models were built by using various sequence-based feature sets and their combinations for enhancer prediction. The selected features derived from a recursive feature elimination method showed that the model using a combination of 141 transcription factor binding motif occurrences from 1,422 transcription factor position weight matrices achieved a favorably high prediction accuracy superior to that of other reported methods. The models demonstrated good prediction accuracy for different enhancer datasets obtained from different cell lines/tissues. In addition, prediction accuracy was further improved by integration of chromatin state features. Our method is complementary to wet-lab experimental methods and provides an additional method to identify enhancers.

OMPcontact: An Outer Membrane Protein Inter-Barrel Residue Contact Prediction Method

In the two transmembrane protein types, outer membrane proteins (OMPs) perform diverse important biochemical functions, including substrate transport and passive nutrient uptake and intake. Hence their 3D structures are expected to reveal these functions. Because experimental structures are scarce, predicted 3D structures are more adapted to OMP research instead, and the inter-barrel residue contact is becoming one of the most remarkable features, improving prediction accuracy by describing the structural information of OMPs. To predict OMP structures accurately, we explored an OMP inter-barrel residue contact prediction method: OMPcontact. Multiple OMP-specific features were integrated in the method, including residue evolutionary covariation, topology-based transmembrane segment relative residue position, OMP lipid layer accessibility, and residue evolution conservation. These features describe the properties of a residue pair in different respects: sequential, structural, evolutionary, and biochemical. Within a 3-residues slide window, a Support Vector Machine (SVM) could accurately determinate the inter-barrel contact residue pair using above features. A 5-fold cross-valuation process was applied in testing the OMPcontact performance against a non-redundant OMP set with 75 samples inside. The tests compared four evolutionary covariation methods and screen analyzed the adaptive ones for inter-barrel contact prediction. The results showed our method not only efficiently realized the prediction, but also scored the possibility for residue pairs reliably. This is expected to improve OMP tertiary structure prediction. Therefore, OMPcontact will be helpful in compiling a structural census of outer membrane protein.

R2C: improving ab initio residue contact map prediction using dynamic fusion strategy and Gaussian noise filter

MOTIVATION

Inter-residue contacts in proteins dictate the topology of protein structures. They are crucial for protein folding and structural stability. Accurate prediction of residue contacts especially for long-range contacts is important to the quality of ab inito structure modeling since they can enforce strong restraints to structure assembly.

RESULTS

In this paper, we present a new Residue-Residue Contact predictor called R2C that combines machine learning-based and correlated mutation analysis-based methods, together with a two-dimensional Gaussian noise filter to enhance the long-range residue contact prediction. Our results show that the outputs from the machine learning-based method are concentrated with better performance on short-range contacts; while for correlated mutation analysis-based approach, the predictions are widespread with higher accuracy on long-range contacts. An effective query-driven dynamic fusion strategy proposed here takes full advantages of the two different methods, resulting in an impressive overall accuracy improvement. We also show that the contact map directly from the prediction model contains the interesting Gaussian noise, which has not been discovered before. Different from recent studies that tried to further enhance the quality of contact map by removing its transitive noise, we designed a new two-dimensional Gaussian noise filter, which was especially helpful for reinforcing the long-range residue contact prediction. Tested on recent CASP10/11 datasets, the overall top L/5 accuracy of our final R2C predictor is 17.6%/15.5% higher than the pure machine learning-based method and 7.8%/8.3% higher than the correlated mutation analysis-based approach for the long-range residue contact prediction.

AVAILABILITY AND IMPLEMENTATION

http://www.csbio.sjtu.edu.cn/bioinf/R2C/Contact:hbshen@sjtu.edu.cn

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Protein Residue Contacts and Prediction Methods

In the field of computational structural proteomics, contact predictions have shown new prospects of solving the longstanding problem of ab initio protein structure prediction. In the last few years, application of deep learning algorithms and availability of large protein sequence databases, combined with improvement in methods that derive contacts from multiple sequence alignments, have shown a huge increase in the precision of contact prediction. In addition, these predicted contacts have also been used to build three-dimensional models from scratch.In this chapter, we briefly discuss many elements of protein residue-residue contacts and the methods available for prediction, focusing on a state-of-the-art contact prediction tool, DNcon. Illustrating with a case study, we describe how DNcon can be used to make ab initio contact predictions for a given protein sequence and discuss how the predicted contacts may be analyzed and evaluated.

A new ensemble coevolution system for detecting HIV-1 protein coevolution

BACKGROUND

A key challenge in the field of HIV-1 protein evolution is the identification of coevolving amino acids at the molecular level. In the past decades, many sequence-based methods have been designed to detect position-specific coevolution within and between different proteins. However, an ensemble coevolution system that integrates different methods to improve the detection of HIV-1 protein coevolution has not been developed.

RESULTS

We integrated 27 sequence-based prediction methods published between 2004 and 2013 into an ensemble coevolution system. This system allowed combinations of different sequence-based methods for coevolution predictions. Using HIV-1 protein structures and experimental data, we evaluated the performance of individual and combined sequence-based methods in the prediction of HIV-1 intra- and inter-protein coevolution. We showed that sequence-based methods clustered according to their methodology, and a combination of four methods outperformed any of the 27 individual methods. This four-method combination estimated that HIV-1 intra-protein coevolving positions were mainly located in functional domains and physically contacted with each other in the protein tertiary structures. In the analysis of HIV-1 inter-protein coevolving positions between Gag and protease, protease drug resistance positions near the active site mostly coevolved with Gag cleavage positions (V128, S373-T375, A431, F448-P453) and Gag C-terminal positions (S489-Q500) under selective pressure of protease inhibitors.

CONCLUSIONS

This study presents a new ensemble coevolution system which detects position-specific coevolution using combinations of 27 different sequence-based methods. Our findings highlight key coevolving residues within HIV-1 structural proteins and between Gag and protease, shedding light on HIV-1 intra- and inter-protein coevolution.

Combining physicochemical and evolutionary information for protein contact prediction

We introduce a novel contact prediction method that achieves high prediction accuracy by combining evolutionary and physicochemical information about native contacts. We obtain evolutionary information from multiple-sequence alignments and physicochemical information from predicted ab initio protein structures. These structures represent low-energy states in an energy landscape and thus capture the physicochemical information encoded in the energy function. Such low-energy structures are likely to contain native contacts, even if their overall fold is not native. To differentiate native from non-native contacts in those structures, we develop a graph-based representation of the structural context of contacts. We then use this representation to train an support vector machine classifier to identify most likely native contacts in otherwise non-native structures. The resulting contact predictions are highly accurate. As a result of combining two sources of information--evolutionary and physicochemical--we maintain prediction accuracy even when only few sequence homologs are present. We show that the predicted contacts help to improve ab initio structure prediction. A web service is available at http://compbio.robotics.tu-berlin.de/epc-map/.

MutPred Splice: machine learning-based prediction of exonic variants that disrupt splicing

We have developed a novel machine-learning approach, MutPred Splice, for the identification of coding region substitutions that disrupt pre-mRNA splicing. Applying MutPred Splice to human disease-causing exonic mutations suggests that 16% of mutations causing inherited disease and 10 to 14% of somatic mutations in cancer may disrupt pre-mRNA splicing. For inherited disease, the main mechanism responsible for the splicing defect is splice site loss, whereas for cancer the predominant mechanism of splicing disruption is predicted to be exon skipping via loss of exonic splicing enhancers or gain of exonic splicing silencer elements. MutPred Splice is available at http://mutdb.org/mutpredsplice.

Sequence-based Gaussian network model for protein dynamics

Motivation: Gaussian network model (GNM) is widely adopted to analyze and understand protein dynamics, function and conformational changes. The existing GNM-based approaches require atomic coordinates of the corresponding protein and cannot be used when only the sequence is known.

Results: We report, first of its kind, GNM model that allows modeling using the sequence. Our linear regression-based, parameter-free, sequence-derived GNM (L-pfSeqGNM) uses contact maps predicted from the sequence and models local, in the sequence, contact neighborhoods with the linear regression. Empirical benchmarking shows relatively high correlations between the native and the predicted with L-pfSeqGNM B-factors and between the cross-correlations of residue fluctuations derived from the structure- and the sequence-based GNM models. Our results demonstrate that L-pfSeqGNM is an attractive platform to explore protein dynamics. In contrast to the highly used GNMs that require protein structures that number in thousands, our model can be used to study motions for the millions of the readily available sequences, which finds applications in modeling conformational changes, protein–protein interactions and protein functions.

Contact:zerozhua@126.com

Supplementary information:Supplementary data are available at Bioinformatics online.

A study and benchmark of DNcon: a method for protein residue-residue contact prediction using deep networks

BACKGROUND

In recent years, the use and importance of predicted protein residue-residue contacts has grown considerably with demonstrated applications such as drug design, protein tertiary structure prediction and model quality assessment. Nevertheless, reported accuracies in the range of 25-35% stubbornly remain the norm for sequence based, long range contact predictions on hard targets. This is in spite of a prolonged effort on behalf of the community to improve the performance of residue-residue contact prediction. A thorough study of the quality of current residue-residue contact predictions and the evaluation metrics used as well as an analysis of current methods is needed to stimulate further advancement in contact prediction and its application. Such a study will better explain the quality and nature of residue-residue contact predictions generated by current methods and as a result lead to better use of this contact information.

RESULTS

We evaluated several sequence based residue-residue contact predictors that participated in the tenth Critical Assessment of protein Structure Prediction (CASP) experiment. The evaluation was performed using standard assessment techniques such as those used by the official CASP assessors as well as two novel evaluation metrics (i.e., cluster accuracy and cluster count). An in-depth analysis revealed that while most residue-residue contact predictions generated are not accurate at the residue level, there is quite a strong contact signal present when allowing for less than residue level precision. Our residue-residue contact predictor, DNcon, performed particularly well achieving an accuracy of 66% for the top L/10 long range contacts when evaluated in a neighbourhood of size 2. The coverage of residue-residue contact areas was also greater with DNcon when compared to other methods. We also provide an analysis of DNcon with respect to its underlying architecture and features used for classification.

CONCLUSIONS

Our novel evaluation metrics demonstrate that current residue-residue contact predictions do contain a strong contact signal and are of better quality than standard evaluation metrics indicate. Our method, DNcon, is a robust, state-of-the-art residue-residue sequence based contact predictor and excelled under a number of evaluation schemes. It is available as a web service at http://iris.rnet.missouri.edu/dncon/.

Evaluation of residue-residue contact prediction in CASP10

We present the results of the assessment of the intramolecular residue-residue contact predictions from 26 prediction groups participating in the 10th round of the CASP experiment. The most recently developed direct coupling analysis methods did not take part in the experiment likely because they require a very deep sequence alignment not available for any of the 114 CASP10 targets. The performance of contact prediction methods was evaluated with the measures used in previous CASPs (i.e., prediction accuracy and the difference between the distribution of the predicted contacts and that of all pairs of residues in the target protein), as well as new measures, such as the Matthews correlation coefficient, the area under the precision-recall curve and the ranks of the first correctly and incorrectly predicted contact. We also evaluated the ability to detect interdomain contacts and tested whether the difficulty of predicting contacts depends upon the protein length and the depth of the family sequence alignment. The analyses were carried out on the target domains for which structural homologs did not exist or were difficult to identify. The evaluation was performed for all types of contacts (short, medium, and long-range), with emphasis placed on long-range contacts, i.e. those involving residues separated by at least 24 residues along the sequence. The assessment suggests that the best CASP10 contact prediction methods perform at approximately the same level, and comparably to those participating in CASP9.

Discrimination of soluble and aggregation-prone proteins based on sequence information

Understanding the factors governing protein solubility is a key to grasp the mechanisms of protein solubility and may provide insight into protein aggregation and misfolding related diseases such as Alzheimer's disease. In this work, we attempt to identify factors important to protein solubility using feature selection. Firstly, we calculate 1438 features including physicochemical properties and statistics for each protein. Random Forest algorithm is used to select the most informative and the minimal subset of features based on their predictive performance. A predictive model is built based on 17 selected features. Compared with previous models, our model achieves better performance with a sensitivity of 0.82, specificity 0.85, ACC 0.84, AUC 0.91 and MCC 0.67. Furthermore, a model using a redundancy-reduced dataset (sequence identity <= 30%) achieves the same performance as the model without redundancy reduction. Our results provide not only a reliable model for predicting protein solubility but also a list of features important to protein solubility. The predictive model is implemented as a freely available web application at .

An integrative computational framework based on a two-step random forest algorithm improves prediction of zinc-binding sites in proteins

Zinc-binding proteins are the most abundant metalloproteins in the Protein Data Bank where the zinc ions usually have catalytic, regulatory or structural roles critical for the function of the protein. Accurate prediction of zinc-binding sites is not only useful for the inference of protein function but also important for the prediction of 3D structure. Here, we present a new integrative framework that combines multiple sequence and structural properties and graph-theoretic network features, followed by an efficient feature selection to improve prediction of zinc-binding sites. We investigate what information can be retrieved from the sequence, structure and network levels that is relevant to zinc-binding site prediction. We perform a two-step feature selection using random forest to remove redundant features and quantify the relative importance of the retrieved features. Benchmarking on a high-quality structural dataset containing 1,103 protein chains and 484 zinc-binding residues, our method achieved >80% recall at a precision of 75% for the zinc-binding residues Cys, His, Glu and Asp on 5-fold cross-validation tests, which is a 10%-28% higher recall at the 75% equal precision compared to SitePredict and zincfinder at residue level using the same dataset. The independent test also indicates that our method has achieved recall of 0.790 and 0.759 at residue and protein levels, respectively, which is a performance better than the other two methods. Moreover, AUC (the Area Under the Curve) and AURPC (the Area Under the Recall-Precision Curve) by our method are also respectively better than those of the other two methods. Our method can not only be applied to large-scale identification of zinc-binding sites when structural information of the target is available, but also give valuable insights into important features arising from different levels that collectively characterize the zinc-binding sites. The scripts and datasets are available at http://protein.cau.edu.cn/zincidentifier/.

Effective inter-residue contact definitions for accurate protein fold recognition

BACKGROUND

Effective encoding of residue contact information is crucial for protein structure prediction since it has a unique role to capture long-range residue interactions compared to other commonly used scoring terms. The residue contact information can be incorporated in structure prediction in several different ways: It can be incorporated as statistical potentials or it can be also used as constraints in ab initio structure prediction. To seek the most effective definition of residue contacts for template-based protein structure prediction, we evaluated 45 different contact definitions, varying bases of contacts and distance cutoffs, in terms of their ability to identify proteins of the same fold.

RESULTS

We found that overall the residue contact pattern can distinguish protein folds best when contacts are defined for residue pairs whose Cβ atoms are at 7.0 Å or closer to each other. Lower fold recognition accuracy was observed when inaccurate threading alignments were used to identify common residue contacts between protein pairs. In the case of threading, alignment accuracy strongly influences the fraction of common contacts identified among proteins of the same fold, which eventually affects the fold recognition accuracy. The largest deterioration of the fold recognition was observed for β-class proteins when the threading methods were used because the average alignment accuracy was worst for this fold class. When results of fold recognition were examined for individual proteins, we found that the effective contact definition depends on the fold of the proteins. A larger distance cutoff is often advantageous for capturing spatial arrangement of the secondary structures which are not physically in contact. For capturing contacts between neighboring β strands, considering the distance between Cα atoms is better than the Cβ-based distance because the side-chain of interacting residues on β strands sometimes point to opposite directions.

CONCLUSION

Residue contacts defined by Cβ-Cβ distance of 7.0 Å work best overall among tested to identify proteins of the same fold. We also found that effective contact definitions differ from fold to fold, suggesting that using different residue contact definition specific for each template will lead to improvement of the performance of threading.

PROTS-RF: a robust model for predicting mutation-induced protein stability changes

The ability to improve protein thermostability via protein engineering is of great scientific interest and also has significant practical value. In this report we present PROTS-RF, a robust model based on the Random Forest algorithm capable of predicting thermostability changes induced by not only single-, but also double- or multiple-point mutations. The model is built using 41 features including evolutionary information, secondary structure, solvent accessibility and a set of fragment-based features. It achieves accuracies of 0.799,0.782, 0.787, and areas under receiver operating characteristic (ROC) curves of 0.873, 0.868 and 0.862 for single-, double- and multiple- point mutation datasets, respectively. Contrary to previous suggestions, our results clearly demonstrate that a robust predictive model trained for predicting single point mutation induced thermostability changes can be capable of predicting double and multiple point mutations. It also shows high levels of robustness in the tests using hypothetical reverse mutations. We demonstrate that testing datasets created based on physical principles can be highly useful for testing the robustness of predictive models.

Predicting protein residue-residue contacts using deep networks and boosting

MOTIVATION

Protein residue-residue contacts continue to play a larger and larger role in protein tertiary structure modeling and evaluation. Yet, while the importance of contact information increases, the performance of sequence-based contact predictors has improved slowly. New approaches and methods are needed to spur further development and progress in the field.

RESULTS

Here we present DNCON, a new sequence-based residue-residue contact predictor using deep networks and boosting techniques. Making use of graphical processing units and CUDA parallel computing technology, we are able to train large boosted ensembles of residue-residue contact predictors achieving state-of-the-art performance.

AVAILABILITY

The web server of the prediction method (DNCON) is available at http://iris.rnet.missouri.edu/dncon/.

CONTACT

chengji@missouri.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Contact map prediction using a large-scale ensemble of rule sets and the fusion of multiple predicted structural features

MOTIVATION

The prediction of a protein's contact map has become in recent years, a crucial stepping stone for the prediction of the complete 3D structure of a protein. In this article, we describe a methodology for this problem that was shown to be successful in CASP8 and CASP9. The methodology is based on (i) the fusion of the prediction of a variety of structural aspects of protein residues, (ii) an ensemble strategy used to facilitate the training process and (iii) a rule-based machine learning system from which we can extract human-readable explanations of the predictor and derive useful information about the contact map representation.

RESULTS

The main part of the evaluation is the comparison against the sequence-based contact prediction methods from CASP9, where our method presented the best rank in five out of the six evaluated metrics. We also assess the impact of the size of the ensemble used in our predictor to show the trade-off between performance and training time of our method. Finally, we also study the rule sets generated by our machine learning system. From this analysis, we are able to estimate the contribution of the attributes in our representation and how these interact to derive contact predictions.

AVAILABILITY

http://icos.cs.nott.ac.uk/servers/psp.html.

CONTACT

natalio.krasnogor@nottingham.ac.uk

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Data mining in the Life Sciences with Random Forest: a walk in the park or lost in the jungle?

In the Life Sciences 'omics' data is increasingly generated by different high-throughput technologies. Often only the integration of these data allows uncovering biological insights that can be experimentally validated or mechanistically modelled, i.e. sophisticated computational approaches are required to extract the complex non-linear trends present in omics data. Classification techniques allow training a model based on variables (e.g. SNPs in genetic association studies) to separate different classes (e.g. healthy subjects versus patients). Random Forest (RF) is a versatile classification algorithm suited for the analysis of these large data sets. In the Life Sciences, RF is popular because RF classification models have a high-prediction accuracy and provide information on importance of variables for classification. For omics data, variables or conditional relations between variables are typically important for a subset of samples of the same class. For example: within a class of cancer patients certain SNP combinations may be important for a subset of patients that have a specific subtype of cancer, but not important for a different subset of patients. These conditional relationships can in principle be uncovered from the data with RF as these are implicitly taken into account by the algorithm during the creation of the classification model. This review details some of the to the best of our knowledge rarely or never used RF properties that allow maximizing the biological insights that can be extracted from complex omics data sets using RF.

The MULTICOM toolbox for protein structure prediction

Background

As genome sequencing is becoming routine in biomedical research, the total number of protein sequences is increasing exponentially, recently reaching over 108 million. However, only a tiny portion of these proteins (i.e. ~75,000 or < 0.07%) have solved tertiary structures determined by experimental techniques. The gap between protein sequence and structure continues to enlarge rapidly as the throughput of genome sequencing techniques is much higher than that of protein structure determination techniques. Computational software tools for predicting protein structure and structural features from protein sequences are crucial to make use of this vast repository of protein resources.

Results

To meet the need, we have developed a comprehensive MULTICOM toolbox consisting of a set of protein structure and structural feature prediction tools. These tools include secondary structure prediction, solvent accessibility prediction, disorder region prediction, domain boundary prediction, contact map prediction, disulfide bond prediction, beta-sheet topology prediction, fold recognition, multiple template combination and alignment, template-based tertiary structure modeling, protein model quality assessment, and mutation stability prediction.

Conclusions

These tools have been rigorously tested by many users in the last several years and/or during the last three rounds of the Critical Assessment of Techniques for Protein Structure Prediction (CASP7-9) from 2006 to 2010, achieving state-of-the-art or near performance. In order to facilitate bioinformatics research and technological development in the field, we have made the MULTICOM toolbox freely available as web services and/or software packages for academic use and scientific research. It is available at http://sysbio.rnet.missouri.edu/multicom_toolbox/.