Learning Proteome Domain Folding Using LSTMs in an Empirical Kernel Space

The recognition of protein structural folds is the starting point for protein function inference and for many structural prediction tools. We previously introduced the idea of using empirical comparisons to create a data-augmented feature space called PESS (Protein Empirical Structure Space)¹ as a novel approach for protein structure prediction. Here, we extend the previous approach by generating the PESS feature space over fixed-length subsequences of query peptides, and applying a sequential neural network model, with one long short-term memory cell layer followed by a fully connected layer. Using this approach, we show that only a small group of domains as a training set is needed to achieve near state-of-the-art accuracy on fold recognition. Our method improves on the previous approach by reducing the training set required and improving the model’s ability to generalize across species, which will help fold prediction for newly discovered proteins.     

GLASS: A tool to visualize protein structure prediction data in three dimensions and evaluate their consistency

When a protein sequence does not share any significant sequence similarity with a protein of known structure, homology modeling cannot be applied. However, many novel and interesting methods, such as secondary structure prediction, fold recognition, and prediction of long-range interactions, are being developed and have been shown to be reasonably successful in predicting protein structures from sequence data and evolutionary information. The a priori evaluation of the correctness of a prediction obtained by one of these methods is however often problematic. Consequently, it is important to use all available information provided by as many different methods as possible and all the available experimental data about the protein of interest, since the consistency of the results is indicative of the reliability of the prediction. Hence the need has arisen for suitable tools able to compare results provided by different methods and evaluate their consistency. We have therefore constructed GLASS, a general platform to read, visualize, compare, and evaluate prediction results from many different sources and to project these prediction results into three dimensions. In addition, GLASS allows the comparison of selected parameters calculated for a model with the distribution observed in real protein structures, thus providing an easy way to test new methods for evaluating the likelihood of different structural models. GLASS can be considered as a “workbench” for structural predictions useful to both experimentalists and theoreticians. Proteins 30:339–351, 1998. © 1998 Wiley-Liss, Inc.     

Bioinformatics methods to predict protein structure and function. A practical approach

Protein structure prediction by using bioinformatics can involve sequence similarity searches, multiple sequence alignments, identification and characterization of domains, secondary structure prediction, solvent accessibility prediction, automatic protein fold recognition, constructing three-dimensional models to atomic detail, and model validation. Not all protein structure prediction projects involve the use of all these techniques. A central part of a typical protein structure prediction is the identification of a suitable structural target from which to extrapolate three-dimensional information for a query sequence. The way in which this is done defines three types of projects. The first involves the use of standard and well-understood techniques. If a structural template remains elusive, a second approach using nontrivial methods is required. If a target fold cannot be reliably identified because inconsistent results have been obtained from nontrivial data analyses, the project falls into the third type of project and will be virtually impossible to complete with any degree of reliability. In this article, a set of protocols to predict protein structure from sequence is presented and distinctions among the three types of project are given. These methods, if used appropriately, can provide valuable indicators of protein structure and function.     

Protein structure prediction using a docking-based hierarchical folding scheme

The pathways by which proteins fold into their specific native structure are still an unsolved mystery. Currently, many methods for protein structure prediction are available, and most of them tackle the problem by relying on the vast amounts of data collected from known protein structures. These methods are often not concerned with the route the protein follows to reach its final fold. This work is based on the premise that proteins fold in a hierarchical manner. We present FOBIA, an automated method for predicting a protein structure. FOBIA consists of two main stages: the first finds matches between parts of the target sequence and independently folding structural units using profile-profile comparison. The second assembles these units into a 3D structure by searching and ranking their possible orientations toward each other using a docking-based approach. We have previously reported an application of an initial version of this strategy to homology based targets. Since then we have considerably enhanced our method's abilities to allow it to address the more difficult template-based target category. This allows us to now apply FOBIA to the template-based targets of CASP8 and to show that it is both very efficient and promising. Our method can provide an alternative for template-based structure prediction, and in particular, the docking-basedranking technique presented here can be incorporated into any profile-profile comparison based method.     

Predicting helix orientation for coiled-coil dimers

The alpha-helical coiled coil is a structurally simple protein oligomerization or interaction motif consisting of two or more alpha helices twisted into a supercoiled bundle. Coiled coils can differ in their stoichiometry, helix orientation, and axial alignment. Because of the near degeneracy of many of these variants, coiled coils pose a challenge to fold recognition methods for structure prediction. Whereas distinctions between some protein folds can be discriminated on the basis of hydrophobic/polar patterning or secondary structure propensities, the sequence differences that encode important details of coiled-coil structure can be subtle. This is emblematic of a larger problem in the field of protein structure and interaction prediction: that of establishing specificity between closely similar structures. We tested the behavior of different computational models on the problem of recognizing the correct orientation--parallel vs. antiparallel--of pairs of alpha helices that can form a dimeric coiled coil. For each of 131 examples of known structure, we constructed a large number of both parallel and antiparallel structural models and used these to assess the ability of five energy functions to recognize the correct fold. We also developed and tested three sequence-based approaches that make use of varying degrees of implicit structural information. The best structural methods performed similarly to the best sequence methods, correctly categorizing approximately 81% of dimers. Steric compatibility with the fold was important for some coiled coils we investigated. For many examples, the correct orientation was determined by smaller energy differences between parallel and antiparallel structures distributed over many residues and energy components. Prediction methods that used structure but incorporated varying approximations and assumptions showed quite different behaviors when used to investigate energetic contributions to orientation preference. Sequence based methods were sensitive to the choice of residue-pair interactions scored.     

Protein folding: a problem with multiple solutions

There is continued interest in predicting the structure of proteins either at the simplest level of identifying their fold class or persevering all the way to an atomic resolution structure. Protein folding methods have become very sophisticated and many successes have been recorded with claims to have solved the native structure of the protein. But for any given protein, there may be more than one solution. Many proteins can exist in one of the other two (or more) different forms and some populate multiple metastable states. Here, the two-state case is considered and the key structural changes that take place when the protein switches from one state to the other are identified. Analysis of these results show that hydrogen bonding patterns and hydrophobic contacts vary considerably between different conformers. Contrary to what has often been assumed previously, these two types of interaction operate essentially independently of one another. Core packing is critical for proper protein structure and function and it is shown that there are considerable changes in internal cavity volumes in many cases. The way in which these switches are made is fold dependent. Considerations such as these need to be taken into account in protein structure prediction.     

Multi-class protein fold recognition using support vector machines and neural networks

MOTIVATION: Protein fold recognition is an important approach to structure discovery without relying on sequence similarity. We study this approach with new multi-class classification methods and examined many issues important for a practical recognition system. RESULTS: Most current discriminative methods for protein fold prediction use the one-against-others method, which has the well-known 'False Positives' problem. We investigated two new methods: the unique one-against-others and the all-against-all methods. Both improve prediction accuracy by 14-110% on a dataset containing 27 SCOP folds. We used the Support Vector Machine (SVM) and the Neural Network (NN) learning methods as base classifiers. SVMs converges fast and leads to high accuracy. When scores of multiple parameter datasets are combined, majority voting reduces noise and increases recognition accuracy. We examined many issues involved with large number of classes, including dependencies of prediction accuracy on the number of folds and on the number of representatives in a fold. Overall, recognition systems achieve 56% fold prediction accuracy on a protein test dataset, where most of the proteins have below 25% sequence identity with the proteins used in training.     

A machine learning information retrieval approach to protein fold recognition

MOTIVATION: Recognizing proteins that have similar tertiary structure is the key step of template-based protein structure prediction methods. Traditionally, a variety of alignment methods are used to identify similar folds, based on sequence similarity and sequence-structure compatibility. Although these methods are complementary, their integration has not been thoroughly exploited. Statistical machine learning methods provide tools for integrating multiple features, but so far these methods have been used primarily for protein and fold classification, rather than addressing the retrieval problem of fold recognition-finding a proper template for a given query protein. RESULTS: Here we present a two-stage machine learning, information retrieval, approach to fold recognition. First, we use alignment methods to derive pairwise similarity features for query-template protein pairs. We also use global profile-profile alignments in combination with predicted secondary structure, relative solvent accessibility, contact map and beta-strand pairing to extract pairwise structural compatibility features. Second, we apply support vector machines to these features to predict the structural relevance (i.e. in the same fold or not) of the query-template pairs. For each query, the continuous relevance scores are used to rank the templates. The FOLDpro approach is modular, scalable and effective. Compared with 11 other fold recognition methods, FOLDpro yields the best results in almost all standard categories on a comprehensive benchmark dataset. Using predictions of the top-ranked template, the sensitivity is approximately 85, 56, and 27% at the family, superfamily and fold levels respectively. Using the 5 top-ranked templates, the sensitivity increases to 90, 70, and 48%.     

Searching for functional sites in protein structures

An ability to assign protein function from protein structure is important for structural genomics consortia. The complex relationship between protein fold and function highlights the necessity of looking beyond the global fold of a protein to specific functional sites. Many computational methods have been developed that address this issue. These include evolutionary trace methods, methods that involve the calculation and assessment of maximal superpositions, methods based on graph theory, and methods that apply machine learning techniques. Such function prediction techniques have been applied to the identification of enzyme catalytic triads and DNA-binding motifs.     

蛋白质结构预测进展

蛋白质的序列决定结构,结构决定功能。新一代准确的蛋白质结构预测工具为结构生物学、结构生物信息学、药物研发和生命科学等许多领域带来了全新的机遇与挑战,单链蛋白质结构预测的准确率达到与试验方法相媲美的水平。本综述概述了蛋白质结构预测领域的理论基础、发展历程与最新进展,讨论了大量预测的蛋白质结构和基于人工智能的方法如何影响实验结构生物学,最后,分析了当前蛋白质结构预测领域仍未解决的问题以及未来的研究方向。     

FORESST: fold recognition from secondary structure predictions of proteins

MOTIVATION: A method for recognizing the three-dimensional fold from the protein amino acid sequence based on a combination of hidden Markov models (HMMs) and secondary structure prediction was recently developed for proteins in the Mainly-Alpha structural class. Here, this methodology is extended to Mainly-Beta and Alpha-Beta class proteins. Compared to other fold recognition methods based on HMMs, this approach is novel in that only secondary structure information is used. Each HMM is trained from known secondary structure sequences of proteins having a similar fold. Secondary structure prediction is performed for the amino acid sequence of a query protein. The predicted fold of a query protein is the fold described by the model fitting the predicted sequence the best. RESULTS: After model cross-validation, the success rate on 44 test proteins covering the three structural classes was found to be 59%. On seven fold predictions performed prior to the publication of experimental structure, the success rate was 71%. In conclusion, this approach manages to capture important information about the fold of a protein embedded in the length and arrangement of the predicted helices, strands and coils along the polypeptide chain. When a more extensive library of HMMs representing the universe of known structural families is available (work in progress), the program will allow rapid screening of genomic databases and sequence annotation when fold similarity is not detectable from the amino acid sequence. AVAILABILITY: FORESST web server at http://absalpha.dcrt.nih.gov:8008/ for the library of HMMs of structural families used in this paper. FORESST web server at http://www.tigr.org/ for a more extensive library of HMMs (work in progress). CONTACT: valedf@tigr.org; munson@helix.nih.gov; garnier@helix.nih.gov     

FLORA: A Novel Method to Predict Protein Function from Structure in Diverse Superfamilies

Predicting protein function from structure remains an active area of interest, particularly for the structural genomics initiatives where a substantial number of structures are initially solved with little or no functional characterisation. Although global structure comparison methods can be used to transfer functional annotations, the relationship between fold and function is complex, particularly in functionally diverse superfamilies that have evolved through different secondary structure embellishments to a common structural core. The majority of prediction algorithms employ local templates built on known or predicted functional residues. Here, we present a novel method (FLORA) that automatically generates structural motifs associated with different functional sub-families (FSGs) within functionally diverse domain superfamilies. Templates are created purely on the basis of their specificity for a given FSG, and the method makes no prior prediction of functional sites, nor assumes specific physico-chemical properties of residues. FLORA is able to accurately discriminate between homologous domains with different functions and substantially outperforms (a 2–3 fold increase in coverage at low error rates) popular structure comparison methods and a leading function prediction method. We benchmark FLORA on a large data set of enzyme superfamilies from all three major protein classes (α, β, αβ) and demonstrate the functional relevance of the motifs it identifies. We also provide novel predictions of enzymatic activity for a large number of structures solved by the Protein Structure Initiative. Overall, we show that FLORA is able to effectively detect functionally similar protein domain structures by purely using patterns of structural conservation of all residues.     

Incorporation of local structural preference potential improves fold recognition

Fold recognition, or threading, is a popular protein structure modeling approach that uses known structure templates to build structures for those of unknown. The key to the success of fold recognition methods lies in the proper integration of sequence, physiochemical and structural information. Here we introduce another type of information, local structural preference potentials of 3-residue and 9-residue fragments, for fold recognition. By combining the two local structural preference potentials with the widely used sequence profile, secondary structure information and hydrophobic score, we have developed a new threading method called FR-t5 (fold recognition by use of 5 terms). In benchmark testings, we have found the consideration of local structural preference potentials in FR-t5 not only greatly enhances the alignment accuracy and recognition sensitivity, but also significantly improves the quality of prediction models.     

Structure- and sequence-based function prediction for non-homologous proteins

The structural genomics projects have been accumulating an increasing number of protein structures, many of which remain functionally unknown. In parallel effort to experimental methods, computational methods are expected to make a significant contribution for functional elucidation of such proteins. However, conventional computational methods that transfer functions from homologous proteins do not help much for these uncharacterized protein structures because they do not have apparent structural or sequence similarity with the known proteins. Here, we briefly review two avenues of computational function prediction methods, i.e. structure-based methods and sequence-based methods. The focus is on our recent developments of local structure-based and sequence-based methods, which can effectively extract function information from distantly related proteins. Two structure-based methods, Pocket-Surfer and Patch-Surfer, identify similar known ligand binding sites for pocket regions in a query protein without using global protein fold similarity information. Two sequence-based methods, protein function prediction and extended similarity group, make use of weakly similar sequences that are conventionally discarded in homology based function annotation. Combined together with experimental methods we hope that computational methods will make leading contribution in functional elucidation of the protein structures.     

Support Vector Machine-based classification of protein folds using the structural properties of amino acid residues and amino acid residue pairs

MOTIVATION: Fold recognition is a key step in the protein structure discovery process, especially when traditional sequence comparison methods fail to yield convincing structural homologies. Although many methods have been developed for protein fold recognition, their accuracies remain low. This can be attributed to insufficient exploitation of fold discriminatory features. RESULTS: We have developed a new method for protein fold recognition using structural information of amino acid residues and amino acid residue pairs. Since protein fold recognition can be treated as a protein fold classification problem, we have developed a Support Vector Machine (SVM) based classifier approach that uses secondary structural state and solvent accessibility state frequencies of amino acids and amino acid pairs as feature vectors. Among the individual properties examined secondary structural state frequencies of amino acids gave an overall accuracy of 65.2% for fold discrimination, which is better than the accuracy by any method reported so far in the literature. Combination of secondary structural state frequencies with solvent accessibility state frequencies of amino acids and amino acid pairs further improved the fold discrimination accuracy to more than 70%, which is approximately 8% higher than the best available method. In this study we have also tested, for the first time, an all-together multi-class method known as Crammer and Singer method for protein fold classification. Our studies reveal that the three multi-class classification methods, namely one versus all, one versus one and Crammer and Singer method, yield similar predictions. AVAILABILITY: Dataset and stand-alone program are available upon request.     

蛋白质折叠识别方法综述

蛋白质折叠识别算法是蛋白质三维结构预测的重要方法之一,该方法在生物科学的许多方面得到卓有成效的应用。在过去的十年中,我们见证了一系列基于不同计算方式的蛋白质折叠识别方法。在这些计算方法中,机器学习和序列谱-序列谱比对是两种在蛋白质折叠中应用较为广泛和有效的方法。除了计算方法的进展外,不断增大的蛋白质结构数据库也是蛋白质折叠识别的预测精度不断提高的一个重要因素。在这篇文章中,我们将简要地回顾蛋白质折叠中的先进算法。另外,我们也将讨论一些可能可以应用于改进蛋白质折叠算法的策略。     

Deciphering a novel thioredoxin-like fold family

Sequence--and structure-based searching strategies have proven useful in the identification of remote homologs and have facilitated both structural and functional predictions of many uncharacterized protein families. We implement these strategies to predict the structure of and to classify a previously uncharacterized cluster of orthologs (COG3019) in the thioredoxin-like fold superfamily. The results of each searching method indicate that thioltransferases are the closest structural family to COG3019. We substantiate this conclusion using the ab initio structure prediction method rosetta, which generates a thioredoxin-like fold similar to that of the glutaredoxin-like thioltransferase (NrdH) for a COG3019 target sequence. This structural model contains the thiol-redox functional motif CYS-X-X-CYS in close proximity to other absolutely conserved COG3019 residues, defining a novel thioredoxin-like active site that potentially binds metal ions. Finally, the rosetta-derived model structure assists us in assembling a global multiple-sequence alignment of COG3019 with two other thioredoxin-like fold families, the thioltransferases and the bacterial arsenate reductases (ArsC).     

The recognition of multi-class protein folds by adding average chemical shifts of secondary structure elements

The recognition of protein folds is an important step in the prediction of protein structure and function. Recently, an increasing number of researchers have sought to improve the methods for protein fold recognition. Following the construction of a dataset consisting of 27 protein fold classes by Ding and Dubchak in 2001, prediction algorithms, parameters and the construction of new datasets have improved for the prediction of protein folds. In this study, we reorganized a dataset consisting of 76-fold classes constructed by Liu et al. and used the values of the increment of diversity, average chemical shifts of secondary structure elements and secondary structure motifs as feature parameters in the recognition of multi-class protein folds. With the combined feature vector as the input parameter for the Random Forests algorithm and ensemble classification strategy, we propose a novel method to identify the 76 protein fold classes. The overall accuracy of the test dataset using an independent test was 66.69%; when the training and test sets were combined, with 5-fold cross-validation, the overall accuracy was 73.43%. This method was further used to predict the test dataset and the corresponding structural classification of the first 27-protein fold class dataset, resulting in overall accuracies of 79.66% and 93.40%, respectively. Moreover, when the training set and test sets were combined, the accuracy using 5-fold cross-validation was 81.21%. Additionally, this approach resulted in improved prediction results using the 27-protein fold class dataset constructed by Ding and Dubchak.     

Protein homology detection and fold inference through multiple alignment entropy profiles

Homology detection and protein structure prediction are central themes in bioinformatics. Establishment of relationship between protein sequences or prediction of their structure by sequence comparison methods finds limitations when there is low sequence similarity. Recent works demonstrate that the use of profiles improves homology detection and protein structure prediction. Profiles can be inferred from protein multiple alignments using different approaches. The "Conservatism-of-Conservatism" is an effective profile analysis method to identify structural features between proteins having the same fold but no detectable sequence similarity. The information obtained from protein multiple alignments varies according to the amino acid classification employed to calculate the profile. In this work, we calculated entropy profiles from PSI-BLAST-derived multiple alignments and used different amino acid classifications summarizing almost 500 different attributes. These entropy profiles were converted into pseudocodes which were compared using the FASTA program with an ad-hoc matrix. We tested the performance of our method to identify relationships between proteins with similar fold using a nonredundant subset of sequences having less than 40% of identity. We then compared our results using Coverage Versus Error per query curves, to those obtained by methods like PSI-BLAST, COMPASS and HHSEARCH. Our method, named HIP (Homology Identification with Profiles) presented higher accuracy detecting relationships between proteins with the same fold. The use of different amino acid classifications reflecting a large number of amino acid attributes, improved the recognition of distantly related folds. We propose the use of pseudocodes representing profile information as a fast and powerful tool for homology detection, fold assignment and analysis of evolutionary information enclosed in protein profiles.