A neural network method for prediction of beta-turn types in proteins using evolutionary information

MOTIVATION: The prediction of beta-turns is an important element of protein secondary structure prediction. Recently, a highly accurate neural network based method Betatpred2 has been developed for predicting beta-turns in proteins using position-specific scoring matrices (PSSM) generated by PSI-BLAST and secondary structure information predicted by PSIPRED. However, the major limitation of Betatpred2 is that it predicts only beta-turn and non-beta-turn residues and does not provide any information of different beta-turn types. Thus, there is a need to predict beta-turn types using an approach based on multiple sequence alignment, which will be useful in overall tertiary structure prediction. RESULTS: In the present work, a method has been developed for the prediction of beta-turn types I, II, IV and VIII. For each turn type, two consecutive feed-forward back-propagation networks with a single hidden layer have been used where the first sequence-to-structure network has been trained on single sequences as well as on PSI-BLAST PSSM. The output from the first network along with PSIPRED predicted secondary structure has been used as input for the second-level structure-to-structure network. The networks have been trained and tested on a non-homologous dataset of 426 proteins chains by 7-fold cross-validation. It has been observed that the prediction performance for each turn type is improved significantly by using multiple sequence alignment. The performance has been further improved by using a second level structure-to-structure network and PSIPRED predicted secondary structure information. It has been observed that Type I and II beta-turns have better prediction performance than Type IV and VIII beta-turns. The final network yields an overall accuracy of 74.5, 93.5, 67.9 and 96.5% with MCC values of 0.29, 0.29, 0.23 and 0.02 for Type I, II, IV and VIII beta-turns, respectively, and is better than random prediction. AVAILABILITY: A web server for prediction of beta-turn types I, II, IV and VIII based on above approach is available at http://www.imtech.res.in/raghava/betaturns/ and http://bioinformatics.uams.edu/mirror/betaturns/ (mirror site).     

Role of evolutionary information in prediction of aromatic-backbone NH interactions in proteins

In this study, an attempt has been made to develop a neural network-based method for predicting segments in proteins containing aromatic-backbone NH (Ar-NH) interactions using multiple sequence alignment. We have analyzed 3121 segments seven residues long containing Ar-NH interactions, extracted from 2298 non-redundant protein structures where no two proteins have more than 25% sequence identity. Two consecutive feed-forward neural networks with a single hidden layer have been trained with standard back-propagation as learning algorithm. The performance of the method improves from 0.12 to 0.15 in terms of Matthews correlation coefficient (MCC) value when evolutionary information (multiple alignment obtained from PSI-BLAST) is used as input instead of a single sequence. The performance of the method further improves from MCC 0.15 to 0.20 when secondary structure information predicted by PSIPRED is incorporated in the prediction. The final network yields an overall prediction accuracy of 70.1% and an MCC of 0.20 when tested by five-fold cross-validation. Overall the performance is 15.2% higher than the random prediction. The method consists of two neural networks: (i) a sequence-to-structure network which predicts the aromatic residues involved in Ar-NH interaction from multiple alignment of protein sequences and (ii) a structure-to structure network where the input consists of the output obtained from the first network and predicted secondary structure. Further, the actual position of the donor residue within the 'potential' predicted fragment has been predicted using a separate sequence-to-structure neural network. Based on the present study, a server Ar_NHPred has been developed which predicts Ar-NH interaction in a given amino acid sequence. The web server Ar_NHPred is available at and (mirror site).     

Large-scale prediction of long disordered regions in proteins using random forests

Background  

Many proteins contain disordered regions that lack fixed three-dimensional (3D) structure under physiological conditions but have important biological functions. Prediction of disordered regions in protein sequences is important for understanding protein function and in high-throughput determination of protein structures. Machine learning techniques, including neural networks and support vector machines have been widely used in such predictions. Predictors designed for long disordered regions are usually less successful in predicting short disordered regions. Combining prediction of short and long disordered regions will dramatically increase the complexity of the prediction algorithm and make the predictor unsuitable for large-scale applications. Efficient batch prediction of long disordered regions alone is of greater interest in large-scale proteome studies.     

Disulfide connectivity prediction using recursive neural networks and evolutionary information

MOTIVATION: We focus on the prediction of disulfide bridges in proteins starting from their amino acid sequence and from the knowledge of the disulfide bonding state of each cysteine. The location of disulfide bridges is a structural feature that conveys important information about the protein main chain conformation and can therefore help towards the solution of the folding problem. Existing approaches based on weighted graph matching algorithms do not take advantage of evolutionary information. Recursive neural networks (RNN), on the other hand, can handle in a natural way complex data structures such as graphs whose vertices are labeled by real vectors, allowing us to incorporate multiple alignment profiles in the graphical representation of disulfide connectivity patterns. RESULTS: The core of the method is the use of machine learning tools to rank alternative disulfide connectivity patterns. We develop an ad-hoc RNN architecture for scoring labeled undirected graphs that represent connectivity patterns. In order to compare our algorithm with previous methods, we report experimental results on the SWISS-PROT 39 dataset. We find that using multiple alignment profiles allows us to obtain significant prediction accuracy improvements, clearly demonstrating the important role played by evolutionary information. AVAILABILITY: The Web interface of the predictor is available at http://neural.dsi.unifi.it/cysteines     

Improving the accuracy of transmembrane protein topology prediction using evolutionary information

MOTIVATION: Many important biological processes such as cell signaling, transport of membrane-impermeable molecules, cell-cell communication, cell recognition and cell adhesion are mediated by membrane proteins. Unfortunately, as these proteins are not water soluble, it is extremely hard to experimentally determine their structure. Therefore, improved methods for predicting the structure of these proteins are vital in biological research. In order to improve transmembrane topology prediction, we evaluate the combined use of both integrated signal peptide prediction and evolutionary information in a single algorithm. RESULTS: A new method (MEMSAT3) for predicting transmembrane protein topology from sequence profiles is described and benchmarked with full cross-validation on a standard data set of 184 transmembrane proteins. The method is found to predict both the correct topology and the locations of transmembrane segments for 80% of the test set. This compares with accuracies of 62-72% for other popular methods on the same benchmark. By using a second neural network specifically to discriminate transmembrane from globular proteins, a very low overall false positive rate (0.5%) can also be achieved in detecting transmembrane proteins. AVAILABILITY: An implementation of the described method is available both as a web server (http://www.psipred.net) and as downloadable source code from http://bioinf.cs.ucl.ac.uk/memsat. Both the server and source code files are free to non-commercial users. Benchmark and training data are also available from http://bioinf.cs.ucl.ac.uk/memsat.     

Accurate prediction of interfacial residues in two‐domain proteins using evolutionary information: Implications for three‐dimensional modeling

With the preponderance of multidomain proteins in eukaryotic genomes, it is essential to recognize the constituent domains and their functions. Often function involves communications across the domain interfaces, and the knowledge of the interacting sites is essential to our understanding of the structure–function relationship. Using evolutionary information extracted from homologous domains in at least two diverse domain architectures (single and multidomain), we predict the interface residues corresponding to domains from the two‐domain proteins. We also use information from the three‐dimensional structures of individual domains of two‐domain proteins to train naïve Bayes classifier model to predict the interfacial residues. Our predictions are highly accurate (～85%) and specific (～95%) to the domain–domain interfaces. This method is specific to multidomain proteins which contain domains in at least more than one protein architectural context. Using predicted residues to constrain domain–domain interaction, rigid‐body docking was able to provide us with accurate full‐length protein structures with correct orientation of domains. We believe that these results can be of considerable interest toward rational protein and interaction design, apart from providing us with valuable information on the nature of interactions. Proteins 2014; 82:1219–1234. © 2013 Wiley Periodicals, Inc.     

Sequence based residue depth prediction using evolutionary information and predicted secondary structure

Background  

Residue depth allows determining how deeply a given residue is buried, in contrast to the solvent accessibility that differentiates between buried and solvent-exposed residues. When compared with the solvent accessibility, the depth allows studying deep-level structures and functional sites, and formation of the protein folding nucleus. Accurate prediction of residue depth would provide valuable information for fold recognition, prediction of functional sites, and protein design.     

An SVM method using evolutionary information for the identification of allergenic proteins

This study presents an allergenic protein prediction system that appears to be capable of producing high sensitivity and specificity. The proposed system is based on support vector machine (SVM) using evolutionary information in the form of an amino acid position specific scoring matrix (PSSM). The performance of this system is assessed by a 10-fold cross-validation experiment using a dataset consisting of 693 allergens and 1041 non-allergens obtained from Swiss-Prot and Structural Database of Allergenic Proteins (SDAP). The PSSM method produced an accuracy of 90.1% in comparison to the methods based on SVM using amino acid, dipeptide composition, pseudo (5-tier) amino acid composition that achieved an accuracy of 86.3, 86.5 and 82.1% respectively. The results show that evolutionary information can be useful to build more effective and efficient allergen prediction systems.     

Using protein structural information in evolutionary inference: transmembrane proteins

We present a model of amino acid sequence evolution based on a hidden Markov model that extends to transmembrane proteins previous methods that incorporate protein structural information into phylogenetics. Our model aims to give a better understanding of processes of molecular evolution and to extract structural information from multiple alignments of transmembrane sequences and use such information to improve phylogenetic analyses. This should be of value in phylogenetic studies of transmembrane proteins: for example, mitochondrial proteins have acquired a special importance in phylogenetics and are mostly transmembrane proteins. The improvement in fit to example data sets of our new model relative to less complex models of amino acid sequence evolution is statistically tested. To further illustrate the potential utility of our method, phylogeny estimation is performed on primate CCR5 receptor sequences, sequences of l and m subunits of the light reaction center in purple bacteria, guinea pig sequences with respect to lagomorph and rodent sequences of calcitonin receptor and K-substance receptor, and cetacean sequences of cytochrome b.     

Modeling disordered regions in proteins using Rosetta

Protein structure prediction methods such as Rosetta search for the lowest energy conformation of the polypeptide chain. However, the experimentally observed native state is at a minimum of the free energy, rather than the energy. The neglect of the missing configurational entropy contribution to the free energy can be partially justified by the assumption that the entropies of alternative folded states, while very much less than unfolded states, are not too different from one another, and hence can be to a first approximation neglected when searching for the lowest free energy state. The shortcomings of current structure prediction methods may be due in part to the breakdown of this assumption. Particularly problematic are proteins with significant disordered regions which do not populate single low energy conformations even in the native state. We describe two approaches within the Rosetta structure modeling methodology for treating such regions. The first does not require advance knowledge of the regions likely to be disordered; instead these are identified by minimizing a simple free energy function used previously to model protein folding landscapes and transition states. In this model, residues can be either completely ordered or completely disordered; they are considered disordered if the gain in entropy outweighs the loss of favorable energetic interactions with the rest of the protein chain. The second approach requires identification in advance of the disordered regions either from sequence alone using for example the DISOPRED server or from experimental data such as NMR chemical shifts. During Rosetta structure prediction calculations the disordered regions make only unfavorable repulsive contributions to the total energy. We find that the second approach has greater practical utility and illustrate this with examples from de novo structure prediction, NMR structure calculation, and comparative modeling.     

Deciphering the cause of evolutionary variance within intrinsically disordered regions in human proteins

Why the intrinsically disordered regions evolve within human proteome has became an interesting question for a decade. Till date, it remains an unsolved yet an intriguing issue to investigate why some of the disordered regions evolve rapidly while the rest are highly conserved across mammalian species. Identifying the key biological factors, responsible for the variation in the conservation rate of different disordered regions within the human proteome, may revisit the above issue. We emphasized that among the other biological features (multifunctionality, gene essentiality, protein connectivity, number of unique domains, gene expression level and expression breadth) considered in our study, the number of unique protein domains acts as a strong determinant that negatively influences the conservation of disordered regions. In this context, we justified that proteins having a fewer types of domains preferably need to conserve their disordered regions to enhance their structural flexibility which in turn will facilitate their molecular interactions. In contrast, the selection pressure acting on the stretches of disordered regions is not so strong in the case of multi-domains proteins. Therefore, we reasoned that the presence of conserved disordered stretches may compensate the functions of multiple domains within a single domain protein. Interestingly, we noticed that the influence of the unique domain number and expression level acts differently on the evolution of disordered regions from that of well-structured ones.     

Combining evolutionary and structural information for local protein structure prediction

We study the effects of various factors in representing and combining evolutionary and structural information for local protein structural prediction based on fragment selection. We prepare databases of fragments from a set of non-redundant protein domains. For each fragment, evolutionary information is derived from homologous sequences and represented as estimated effective counts and frequencies of amino acids (evolutionary frequencies) at each position. Position-specific amino acid preferences called structural frequencies are derived from statistical analysis of discrete local structural environments in database structures. Our method for local structure prediction is based on ranking and selecting database fragments that are most similar to a target fragment. Using secondary structure type as a local structural property, we test our method in a number of settings. The major findings are: (1) the COMPASS-type scoring function for fragment similarity comparison gives better prediction accuracy than three other tested scoring functions for profile-profile comparison. We show that the COMPASS-type scoring function can be derived both in the probabilistic framework and in the framework of statistical potentials. (2) Using the evolutionary frequencies of database fragments gives better prediction accuracy than using structural frequencies. (3) Finer definition of local environments, such as including more side-chain solvent accessibility classes and considering the backbone conformations of neighboring residues, gives increasingly better prediction accuracy using structural frequencies. (4) Combining evolutionary and structural frequencies of database fragments, either in a linear fashion or using a pseudocount mixture formula, results in improvement of prediction accuracy. Combination at the log-odds score level is not as effective as combination at the frequency level. This suggests that there might be better ways of combining sequence and structural information than the commonly used linear combination of log-odds scores. Our method of fragment selection and frequency combination gives reasonable results of secondary structure prediction tested on 56 CASP5 targets (average SOV score 0.77), suggesting that it is a valid method for local protein structure prediction. Mixture of predicted structural frequencies and evolutionary frequencies improve the quality of local profile-to-profile alignment by COMPASS.     

Genome‐scale prediction of proteins with long intrinsically disordered regions

Proteins with long disordered regions (LDRs), defined as having 30 or more consecutive disordered residues, are abundant in eukaryotes, and these regions are recognized as a distinct class of biologically functional domains. LDRs facilitate various cellular functions and are important for target selection in structural genomics. Motivated by the lack of methods that directly predict proteins with LDRs, we designed Super‐fast predictor of proteins with Long Intrinsically DisordERed regions (SLIDER). SLIDER utilizes logistic regression that takes an empirically chosen set of numerical features, which consider selected physicochemical properties of amino acids, sequence complexity, and amino acid composition, as its inputs. Empirical tests show that SLIDER offers competitive predictive performance combined with low computational cost. It outperforms, by at least a modest margin, a comprehensive set of modern disorder predictors (that can indirectly predict LDRs) and is 16 times faster compared to the best currently available disorder predictor. Utilizing our time‐efficient predictor, we characterized abundance and functional roles of proteins with LDRs over 110 eukaryotic proteomes. Similar to related studies, we found that eukaryotes have many (on average 30.3%) proteins with LDRs with majority of proteomes having between 25 and 40%, where higher abundance is characteristic to proteomes that have larger proteins. Our first‐of‐its‐kind large‐scale functional analysis shows that these proteins are enriched in a number of cellular functions and processes including certain binding events, regulation of catalytic activities, cellular component organization, biogenesis, biological regulation, and some metabolic and developmental processes. A webserver that implements SLIDER is available at http://biomine.ece.ualberta.ca/SLIDER/ .Proteins 2014; 82:145–158. © 2013 Wiley Periodicals, Inc.     

Delineation of modular proteins: domain boundary prediction from sequence information

The delineation of domain boundaries of a given sequence in the absence of known 3D structures or detectable sequence homology to known domains benefits many areas in protein science, such as protein engineering, protein 3D structure determination and protein structure prediction. With the exponential growth of newly determined sequences, our ability to predict domain boundaries rapidly and accurately from sequence information alone is both essential and critical from the viewpoint of gene function annotation. Anyone attempting to predict domain boundaries for a single protein sequence is invariably confronted with a plethora of databases that contain boundary information available from the internet and a variety of methods for domain boundary prediction. How are these derived and how well do they work? What definition of 'domain' do they use? We will first clarify the different definitions of protein domains, and then describe the available public databases with domain boundary information. Finally, we will review existing domain boundary prediction methods and discuss their strengths and weaknesses.     

Classification of G proteins and prediction of GPCRs-G proteins coupling specificity using continuous wavelet transform and information theory

The coupling between G protein-coupled receptors (GPCRs) and guanine nucleotide-binding proteins (G proteins) regulates various signal transductions from extracellular space into the cell. However, the coupling mechanism between GPCRs and G proteins is still unknown, and experimental determination of their coupling specificity and function is both expensive and time consuming. Therefore, it is significant to develop a theoretical method to predict the coupling specificity between GPCRs and G proteins as well as their function using their primary sequences. In this study, a novel four-layer predictor (GPCRsG_CWTIT) based on support vector machine (SVM), continuous wavelet transform (CWT) and information theory (IT) is developed to classify G proteins and predict the coupling specificity between GPCRs and G proteins. SVM is used for construction of models. CWT and IT are used to characterize the primary structure of protein. Performance of GPCRsG_CWTIT is evaluated with cross-validation test on various working dataset. The overall accuracy of the G proteins at the levels of class and family is 98.23 and 85.42%, respectively. The accuracy of the coupling specificity prediction varies from 74.60 to 94.30%. These results indicate that the proposed predictor is an effective and feasible tool to predict the coupling specificity between GPCRs and G proteins as well as their functions using only the protein full sequence. The establishment of such an accurate prediction method will facilitate drug discovery by improving the ability to identify and predict protein-protein interactions. GPCRsG_CWTIT and dataset can be acquired freely on request from the authors.     

Support Vector Machine-based method for predicting subcellular localization of mycobacterial proteins using evolutionary information and motifs

Background

Annotations that describe the function of sequences are enormously important to researchers during laboratory investigations and when making computational inferences. However, there has been little investigation into the data quality of sequence function annotations. Here we have developed a new method of estimating the error rate of curated sequence annotations, and applied this to the Gene Ontology (GO) sequence database (GOSeqLite). This method involved artificially adding errors to sequence annotations at known rates, and used regression to model the impact on the precision of annotations based on BLAST matched sequences.

Results

We estimated the error rate of curated GO sequence annotations in the GOSeqLite database (March 2006) at between 28% and 30%. Annotations made without use of sequence similarity based methods (non-ISS) had an estimated error rate of between 13% and 18%. Annotations made with the use of sequence similarity methodology (ISS) had an estimated error rate of 49%.

Conclusion

While the overall error rate is reasonably low, it would be prudent to treat all ISS annotations with caution. Electronic annotators that use ISS annotations as the basis of predictions are likely to have higher false prediction rates, and for this reason designers of these systems should consider avoiding ISS annotations where possible. Electronic annotators that use ISS annotations to make predictions should be viewed sceptically. We recommend that curators thoroughly review ISS annotations before accepting them as valid. Overall, users of curated sequence annotations from the GO database should feel assured that they are using a comparatively high quality source of information.     

POTENCI: prediction of temperature,neighbor and pH-corrected chemical shifts for intrinsically disordered proteins

Chemical shifts contain important site-specific information on the structure and dynamics of proteins. Deviations from statistical average values, known as random coil chemical shifts (RCCSs), are extensively used to infer these relationships. Unfortunately, the use of imprecise reference RCCSs leads to biased inference and obstructs the detection of subtle structural features. Here we present a new method, POTENCI, for the prediction of RCCSs that outperforms the currently most authoritative methods. POTENCI is parametrized using a large curated database of chemical shifts for protein segments with validated disorder; It takes pH and temperature explicitly into account, and includes sequence-dependent nearest and next-nearest neighbor corrections as well as second-order corrections. RCCS predictions with POTENCI show root-mean-square values that are lower by 25–78%, with the largest improvements observed for ¹Hα and ¹³C′. It is demonstrated how POTENCI can be applied to analyze subtle deviations from RCCSs to detect small populations of residual structure in intrinsically disorder proteins that were not discernible before. POTENCI source code is available for download, or can be deployed from the URL http://www.protein-nmr.org.     

Better prediction of sub-cellular localization by combining evolutionary and structural information

The native sub-cellular compartment of a protein is one aspect of its function. Thus, predicting localization is an important step toward predicting function. Short zip code-like sequence fragments regulate some of the shuttling between compartments. Cataloguing and predicting such motifs is the most accurate means of determining localization in silico. However, only few motifs are currently known, and not all the trafficking appears regulated in this way. The amino acid composition of a protein correlates with its localization. All general prediction methods employed this observation. Here, we explored the evolutionary information contained in multiple alignments and aspects of protein structure to predict localization in absence of homology and targeting motifs. Our final system combined statistical rules and a variety of neural networks to achieve an overall four-state accuracy above 65%, a significant improvement over systems using only composition. The system was at its best for extra-cellular and nuclear proteins; it was significantly less accurate than TargetP for mitochondrial proteins. Interestingly, all methods that were developed on SWISS-PROT sequences failed grossly when fed with sequences from proteins of known structures taken from PDB. We therefore developed two separate systems: one for proteins of known structure and one for proteins of unknown structure. Finally, we applied the PDB-based system along with homology-based inferences and automatic text analysis to annotate all eukaryotic proteins in the PDB (http://cubic.bioc.columbia.edu/db/LOC3D). We imagine that this pilot method-certainly in combination with similar tools-may be valuable target selection in structural genomics.