Search

Structures of biomolecular systems are increasingly computed by integrative modeling that relies on varied types of experimental data and theoretical information. We describe here the proceedings and conclusions from the first wwPDB Hybrid/Integrative Methods Task Force Workshop held at the European Bioinformatics Institute in Hinxton, UK, on October 6 and 7, 2014. At the workshop, experts in various experimental fields of structural biology, experts in integrative modeling and visualization, and experts in data archiving addressed a series of questions central to the future of structural biology. How should integrative models be represented? How should the data and integrative models be validated? What data should be archived? How should the data and models be archived? What information should accompany the publication of integrative models?, (Copyright © 2015 Elsevier Ltd. All rights reserved.)

Target identification is essential for drug design, drug-drug interaction prediction, dosage adjustment and side effect anticipation. Specifically, the knowledge of structural details is essential for understanding the mode of action of a compound on a target protein. Here, we present nAnnoLyze, a method for target identification that relies on the hypothesis that structurally similar binding sites bind similar ligands. nAnnoLyze integrates structural information into a bipartite network of interactions and similarities to predict structurally detailed compound-protein interactions at proteome scale. The method was benchmarked on a dataset of 6,282 pairs of known interacting ligand-target pairs reaching a 0.96 of area under the Receiver Operating Characteristic curve (AUC) when using the drug names as an input feature for the classifier, and a 0.70 of AUC for "anonymous" compounds or compounds not present in the training set. nAnnoLyze resulted in higher accuracies than its predecessor, AnnoLyze. We applied the method to predict interactions for all the compounds in the DrugBank database with each human protein structure and provide examples of target identification for known drugs against human diseases. The accuracy and applicability of our method to any compound indicate that a comparative docking approach such as nAnnoLyze enables large-scale annotation and analysis of compound-protein interactions and thus may benefit drug development.

Comparing the structures of proteins is crucial to gaining insight into protein evolution and function. Here, we align the sequences of multiple protein structures by a dynamic programming optimization of a scoring function that is a sum of an affine gap penalty and terms dependent on various sequence and structure features (SALIGN). The features include amino acid residue type, residue position, residue accessible surface area, residue secondary structure state and the conformation of a short segment centered on the residue. The multiple alignment is built by following the 'guide' tree constructed from the matrix of all pairwise protein alignment scores. Importantly, the method does not depend on the exact values of various parameters, such as feature weights and gap penalties, because the optimal alignment across a range of parameter values is found. Using multiple structure alignments in the HOMSTRAD database, SALIGN was benchmarked against MUSTANG for multiple alignments as well as against TM-align and CE for pairwise alignments. On the average, SALIGN produces a 15% improvement in structural overlap over HOMSTRAD and 14% over MUSTANG, and yields more equivalent structural positions than TM-align and CE in 90% and 95% of cases, respectively. The utility of accurate multiple structure alignment is illustrated by its application to comparative protein structure modeling.

We introduce a new type of knowledge-based potentials for protein structure prediction, called 'evolutionary potentials', which are derived using a single experimental protein structure and all three-dimensional models of its homologous sequences. The new potentials have been benchmarked against other knowledge-based potentials, resulting in a significant increase in accuracy for model assessment. In contrast to standard knowledge-based potentials, we propose that evolutionary potentials capture key determinants of thermodynamic stability and specific sequence constraints required for fast folding.

Predicting the functional impact of protein variation is one of the most challenging problems in bioinformatics. A rapidly growing number of genome-scale studies provide large amounts of experimental data, allowing the application of rigorous statistical approaches for predicting whether a given single point mutation has an impact on human health. Up until now, existing methods have limited their source data to either protein or gene information. Novel in this work, we take advantage of both and focus on protein evolutionary information by using estimated selective pressures at the codon level. Here we introduce a new method (SeqProfCod) to predict the likelihood that a given protein variant is associated with human disease or not. Our method relies on a support vector machine (SVM) classifier trained using three sources of information: protein sequence, multiple protein sequence alignments, and the estimation of selective pressure at the codon level. SeqProfCod has been benchmarked with a large dataset of 8,987 single point mutations from 1,434 human proteins from SWISS-PROT. It achieves 82% overall accuracy and a correlation coefficient of 0.59, indicating that the estimation of the selective pressure helps in predicting the functional impact of single-point mutations. Moreover, this study demonstrates the synergic effect of combining two sources of information for predicting the functional effects of protein variants: protein sequence/profile-based information and the evolutionary estimation of the selective pressures at the codon level. The results of large-scale application of SeqProfCod over all annotated point mutations in SWISS-PROT (available for download at http://sgu.bioinfo.cipf.es/services/Omidios/; last accessed: 24 August 2007), could be used to support clinical studies., ((c) 2007 Wiley-Liss, Inc.)

The characterization of protein interactions is essential for understanding biological systems. While genome-scale methods are available for identifying interacting proteins, they do not pinpoint the interacting motifs (e.g., a domain, sequence segments, a binding site, or a set of residues). Here, we develop and apply a method for delineating the interacting motifs of hub proteins (i.e., highly connected proteins). The method relies on the observation that proteins with common interaction partners tend to interact with these partners through a common interacting motif. The sole input for the method are binary protein interactions; neither sequence nor structure information is needed. The approach is evaluated by comparing the inferred interacting motifs with domain families defined for 368 proteins in the Structural Classification of Proteins (SCOP). The positive predictive value of the method for detecting proteins with common SCOP families is 75% at sensitivity of 10%. Most of the inferred interacting motifs were significantly associated with sequence patterns, which could be responsible for the common interactions. We find that yeast hubs with multiple interacting motifs are more likely to be essential than hubs with one or two interacting motifs, thus rationalizing the previously observed correlation between essentiality and the number of interacting partners of a protein. We also find that yeast hubs with multiple interacting motifs evolve slower than the average protein, contrary to the hubs with one or two interacting motifs. The proposed method will help us discover unknown interacting motifs and provide biological insights about protein hubs and their roles in interaction networks.

The DBAli tools use a comprehensive set of structural alignments in the DBAli database to leverage the structural information deposited in the Protein Data Bank (PDB). These tools include (i) the DBAlit program that allows users to input the 3D coordinates of a protein structure for comparison by MAMMOTH against all chains in the PDB; (ii) the AnnoLite and AnnoLyze programs that annotate a target structure based on its stored relationships to other structures; (iii) the ModClus program that clusters structures by sequence and structure similarities; (iv) the ModDom program that identifies domains as recurrent structural fragments and (v) an implementation of the COMPARER method in the SALIGN command in MODELLER that creates a multiple structure alignment for a set of related protein structures. Thus, the DBAli tools, which are freely accessible via the World Wide Web at http://salilab.org/DBAli/, allow users to mine the protein structure space by establishing relationships between protein structures and their functions.

Background: Advances in structural biology, including structural genomics, have resulted in a rapid increase in the number of experimentally determined protein structures. However, about half of the structures deposited by the structural genomics consortia have little or no information about their biological function. Therefore, there is a need for tools for automatically and comprehensively annotating the function of protein structures. We aim to provide such tools by applying comparative protein structure annotation that relies on detectable relationships between protein structures to transfer functional annotations. Here we introduce two programs, AnnoLite and AnnoLyze, which use the structural alignments deposited in the DBAli database., Description: AnnoLite predicts the SCOP, CATH, EC, InterPro, PfamA, and GO terms with an average sensitivity of ~90% and average precision of ~80%. AnnoLyze predicts ligand binding site and domain interaction patches with an average sensitivity of ~70% and average precision of ~30%, correctly localizing binding sites for small molecules in ~95% of its predictions., Conclusion: The AnnoLite and AnnoLyze programs for comparative annotation of protein structures can reliably and automatically annotate new protein structures. The programs are fully accessible via the Internet as part of the DBAli suite of tools at http://salilab.org/DBAli/.

The rise in the number of functionally uncharacterized protein structures is increasing the demand for structure-based methods for functional annotation. Here, we describe a method for predicting the location of a binding site of a given type on a target protein structure. The method begins by constructing a scoring function, followed by a Monte Carlo optimization, to find a good scoring patch on the protein surface. The scoring function is a weighted linear combination of the z-scores of various properties of protein structure and sequence, including amino acid residue conservation, compactness, protrusion, convexity, rigidity, hydrophobicity, and charge density; the weights are calculated from a set of previously identified instances of the binding-site type on known protein structures. The scoring function can easily incorporate different types of information useful in localization, thus increasing the applicability and accuracy of the approach. To test the method, 1008 known protein structures were split into 20 different groups according to the type of the bound ligand. For nonsugar ligands, such as various nucleotides, binding sites were correctly identified in 55%-73% of the cases. The method is completely automated (http://salilab.org/patcher) and can be applied on a large scale in a structural genomics setting.

Functional characterization of a protein sequence is one of the most frequent problems in biology. This task is usually facilitated by accurate three-dimensional (3-D) structure of the studied protein. In the absence of an experimentally determined structure, comparative or homology modeling can sometimes provide a useful 3-D model for a protein that is related to at least one known protein structure. Comparative modeling predicts the 3-D structure of a given protein sequence (target) based primarily on its alignment to one or more proteins of known structure (templates). The prediction process consists of fold assignment, target-template alignment, model building, and model evaluation. This unit describes how to calculate comparative models using the program MODELLER and discusses all four steps of comparative modeling, frequently observed errors, and some applications. Modeling lactate dehydrogenase from Trichomonas vaginalis (TvLDH) is described as an example. The download and installation of the MODELLER software is also described.

Reliable prediction of model accuracy is an important unsolved problem in protein structure modeling. To address this problem, we studied 24 individual assessment scores, including physics-based energy functions, statistical potentials, and machine learning-based scoring functions. Individual scores were also used to construct approximately 85,000 composite scoring functions using support vector machine (SVM) regression. The scores were tested for their abilities to identify the most native-like models from a set of 6000 comparative models of 20 representative protein structures. Each of the 20 targets was modeled using a template of <30% sequence identity, corresponding to challenging comparative modeling cases. The best SVM score outperformed all individual scores by decreasing the average RMSD difference between the model identified as the best of the set and the model with the lowest RMSD (DeltaRMSD) from 0.63 A to 0.45 A, while having a higher Pearson correlation coefficient to RMSD (r=0.87) than any other tested score. The most accurate score is based on a combination of the DOPE non-hydrogen atom statistical potential; surface, contact, and combined statistical potentials from MODPIPE; and two PSIPRED/DSSP scores. It was implemented in the SVMod program, which can now be applied to select the final model in various modeling problems, including fold assignment, target-template alignment, and loop modeling.

Reduced or simplified amino acid alphabets group the 20 naturally occurring amino acids into a smaller number of representative protein residues. To date, several reduced amino acid alphabets have been proposed, which have been derived and optimized by a variety of methods. The resulting reduced amino acid alphabets have been applied to pattern recognition, generation of consensus sequences from multiple alignments, protein folding, and protein structure prediction. In this work, amino acid substitution matrices and statistical potentials were derived based on several reduced amino acid alphabets and their performance assessed in a large benchmark for the tasks of sequence alignment and fold assessment of protein structure models, using as a reference frame the standard alphabet of 20 amino acids. The results showed that a large reduction in the total number of residue types does not necessarily translate into a significant loss of discriminative power for sequence alignment and fold assessment. Therefore, some definitions of a few residue types are able to encode most of the relevant sequence/structure information that is present in the 20 standard amino acids. Based on these results, we suggest that the use of reduced amino acid alphabets may allow to increasing the accuracy of current substitution matrices and statistical potentials for the prediction of protein structure of remote homologs., (2006 Wiley-Liss, Inc.)

The penalty for inserting gaps into an alignment between two protein sequences is a major determinant of the alignment accuracy. Here, we present an algorithm for finding a globally optimal alignment by dynamic programming that can use a variable gap penalty (VGP) function of any form. We also describe a specific function that depends on the structural context of an insertion or deletion. It penalizes gaps that are introduced within regions of regular secondary structure, buried regions, straight segments and also between two spatially distant residues. The parameters of the penalty function were optimized on a set of 240 sequence pairs of known structure, spanning the sequence identity range of 20-40%. We then tested the algorithm on another set of 238 sequence pairs of known structures. The use of the VGP function increases the number of correctly aligned residues from 81.0 to 84.5% in comparison with the optimized affine gap penalty function; this difference is statistically significant according to Student's t-test. We estimate that the new algorithm allows us to produce comparative models with an additional approximately 7 million accurately modeled residues in the approximately 1.1 million proteins that are detectably related to a known structure.

The function of an uncharacterized protein is usually inferred either from its homology to, or its interactions with, characterized proteins. Here, we use both sequence similarity and protein interactions to identify relationships between remotely related protein sequences. We rely on the fact that homologous sequences share similar interactions, and, therefore, the set of interacting partners of the partners of a given protein is enriched by its homologs. The approach was bench-marked by assigning the fold and functional family to test sequences of known structure. Specifically, we relied on 1,434 proteins with known folds, as defined in the Structural Classification of Proteins (SCOP) database, and with known interacting partners, as defined in the Database of Interacting Proteins (DIP). For this subset, the specificity of fold assignment was increased from 54% for position-specific iterative BLAST to 75% for our approach, with a concomitant increase in sensitivity for a few percentage points. Similarly, the specificity of family assignment at the e-value threshold of 10(-8) was increased from 70% to 87%. The proposed method would be a useful tool for large-scale automated discovery of remote relationships between protein sequences, given its unique reliance on sequence similarity and protein-protein interactions.

The accuracy of an alignment between two protein sequences can be improved by including other detectably related sequences in the comparison. We optimize and benchmark such an approach that relies on aligning two multiple sequence alignments, each one including one of the two protein sequences. Thirteen different protocols for creating and comparing profiles corresponding to the multiple sequence alignments are implemented in the SALIGN command of MODELLER. A test set of 200 pairwise, structure-based alignments with sequence identities below 40% is used to benchmark the 13 protocols as well as a number of previously described sequence alignment methods, including heuristic pairwise sequence alignment by BLAST, pairwise sequence alignment by global dynamic programming with an affine gap penalty function by the ALIGN command of MODELLER, sequence-profile alignment by PSI-BLAST, Hidden Markov Model methods implemented in SAM and LOBSTER, pairwise sequence alignment relying on predicted local structure by SEA, and multiple sequence alignment by CLUSTALW and COMPASS. The alignment accuracies of the best new protocols were significantly better than those of the other tested methods. For example, the fraction of the correctly aligned residues relative to the structure-based alignment by the best protocol is 56%, which can be compared with the accuracies of 26%, 42%, 43%, 48%, 50%, 49%, 43%, and 43% for the other methods, respectively. The new method is currently applied to large-scale comparative protein structure modeling of all known sequences.

MODBASE (http://salilab.org/modbase) is a relational database of annotated comparative protein structure models for all available protein sequences matched to at least one known protein structure. The models are calculated by MODPIPE, an automated modeling pipeline that relies on the MODELLER package for fold assignment, sequence-structure alignment, model building and model assessment (http:/salilab.org/modeller). MODBASE uses the MySQL relational database management system for flexible querying and CHIMERA for viewing the sequences and structures (http://www.cgl.ucsf.edu/chimera/). MODBASE is updated regularly to reflect the growth in protein sequence and structure databases, as well as improvements in the software for calculating the models. For ease of access, MODBASE is organized into different data sets. The largest data set contains 1,26,629 models for domains in 659,495 out of 1,182,126 unique protein sequences in the complete Swiss-Prot/TrEMBL database (August 25, 2003); only models based on alignments with significant similarity scores and models assessed to have the correct fold despite insignificant alignments are included. Another model data set supports target selection and structure-based annotation by the New York Structural Genomics Research Consortium; e.g. the 53 new structures produced by the consortium allowed us to characterize structurally 24,113 sequences. MODBASE also contains binding site predictions for small ligands and a set of predicted interactions between pairs of modeled sequences from the same genome. Our other resources associated with MODBASE include a comprehensive database of multiple protein structure alignments (DBALI, http://salilab.org/dbali) as well as web servers for automated comparative modeling with MODPIPE (MODWEB, http://salilab. org/modweb), modeling of loops in protein structures (MODLOOP, http://salilab.org/modloop) and predicting functional consequences of single nucleotide polymorphisms (SNPWEB, http://salilab. org/snpweb).

Summary: We describe ModView, a web application for visualization of multiple protein sequences and structures. ModView integrates a multiple structure viewer, a multiple sequence alignment editor, and a database querying engine. It is possible to interactively manipulate hundreds of proteins, to visualize conservative and variable residues, active and binding sites, fragments, and domains in protein families, as well as to display large macromolecular complexes such as ribosomes or viruses. As a Netscape plug-in, ModView can be included in HTML pages along with text and figures, which makes it useful for teaching and presentations. ModView is also suitable as a graphical interface to various databases because it can be controlled through JavaScript commands and called from CGI scripts., Availability: ModView is available at http://guitar.rockefeller.edu/modview.

The reliability of ranking of protein structure modeling methods is assessed. The assessment is based on the parametric Student's t test and the nonparametric Wilcox signed rank test of statistical significance of the difference between paired samples. The approach is applied to the ranking of the comparative modeling methods tested at the fourth meeting on Critical Assessment of Techniques for Protein Structure Prediction (CASP). It is shown that the 14 CASP4 test sequences may not be sufficient to reliably distinguish between the top eight methods, given the model quality differences and their standard deviations. We suggest that CASP needs to be supplemented by an assessment of protein structure prediction methods that is automated, continuous in time, based on several criteria applied to a large number of models, and with quantitative statistical reliability assigned to each characterization.

MODBASE (http://salilab.org/modbase) is a database of annotated comparative protein structure models. The models are calculated by MODPIPE, an automated modeling pipeline that relies primarily on MODELLER for fold assignment, sequence-structure alignment, model building and model assessment (http:/salilab.org/modeller). MODBASE currently contains 5,152,695 reliable models for domains in 1,593,209 unique protein sequences; only models based on statistically significant alignments and/or models assessed to have the correct fold are included. MODBASE also allows users to calculate comparative models on demand, through an interface to the MODWEB modeling server (http://salilab.org/modweb). Other resources integrated with MODBASE include databases of multiple protein structure alignments (DBAli), structurally defined ligand binding sites (LIGBASE), predicted ligand binding sites (AnnoLyze), structurally defined binary domain interfaces (PIBASE) and annotated single nucleotide polymorphisms and somatic mutations found in human proteins (LS-SNP, LS-Mut). MODBASE models are also available through the Protein Model Portal (http://www.proteinmodelportal.org/).

Comparing the structures of proteins is crucial to gaining insight into protein evolution and function. Here, we align the sequences of multiple protein structures by a dynamic programming optimization of a scoring function that is a sum of an affine gap penalty and terms dependent on various sequence and structure features (SALIGN). The features include amino acid residue type, residue position, residue accessible surface area, residue secondary structure state and the conformation of a short segment centered on the residue. The multiple alignment is built by following the ‘guide’ tree constructed from the matrix of all pairwise protein alignment scores. Importantly, the method does not depend on the exact values of various parameters, such as feature weights and gap penalties, because the optimal alignment across a range of parameter values is found. Using multiple structure alignments in the HOMSTRAD database, SALIGN was benchmarked against MUSTANG for multiple alignments as well as against TM-align and CE for pairwise alignments. On the average, SALIGN produces a 15% improvement in structural overlap over HOMSTRAD and 14% over MUSTANG, and yields more equivalent structural positions than TM-align and CE in 90% and 95% of cases, respectively. The utility of accurate multiple structure alignment is illustrated by its application to comparative protein structure modeling. [ABSTRACT FROM PUBLISHER]

Background: A number of studies have used protein interaction data alone for protein function prediction. Here, we introduce a computational approach for annotation of enzymes, based on the observation that similar protein sequences are more likely to perform the same function if they share similar interacting partners. Results: The method has been tested against the PSI-BLAST program using a set of 3,890 protein sequences from which interaction data was available. For protein sequences that align with at least 40% sequence identity to a known enzyme, the specificity of our method in predicting the first three EC digits increased from 80% to 90% at 80% coverage when compared to PSI-BLAST. Conclusion: Our method can also be used in proteins for which homologous sequences with known interacting partners can be detected. Thus, our method could increase 10% the specificity of genome-wide enzyme predictions based on sequence matching by PSI-BLAST alone. [ABSTRACT FROM AUTHOR]

Only few instances are known of protein folds that tolerate massive sequence variation for the sake of binding diversity. The most extensively characterized is the immunoglobulin fold. We now add to this the C-type lectin (CLec) fold, as found in the major tropism determinant (Mtd), a retroelement-encoded receptor-binding protein of Bordetella bacteriophage. Variation in Mtd, with its ∼1013 possible sequences, enables phage adaptation to Bordetella spp. Mtd is an intertwined, pyramid-shaped trimer, with variable residues organized by its CLec fold into discrete receptor-binding sites. The CLec fold provides a highly static scaffold for combinatorial display of variable residues, probably reflecting a different evolutionary solution for balancing diversity against stability from that in the immunoglobulin fold. Mtd variants are biased toward the receptor pertactin, and there is evidence that the CLec fold is used broadly for sequence variation by related retroelements. [ABSTRACT FROM AUTHOR]

Comparative modeling predicts the three-dimensional structure of a given protein sequence (target) based primarily on its alignment to one or more proteins of known structure (templates). The prediction process consists of fold assignment, target–template alignment, model building, and model evaluation. The number of protein sequences that can be modeled and the accuracy of the predictions are increasing steadily because of the growth in the number of known protein structures and because of the improvements in the modeling software. Further advances are necessary in recognizing weak sequence–structure similarities, aligning sequences with structures, modeling of rigid body shifts, distortions, loops and side chains, as well as detecting errors in a model. Despite these problems, it is currently possible to model with useful accuracy significant parts of approximately one third of all known protein sequences. The use of individual comparative models in biology is already rewarding and increasingly widespread. A major new challenge for comparative modeling is the integration of it with the torrents of data from genome sequencing projects as well as from functional and structural genomics. In particular, there is a need to develop an automated, rapid, robust, sensitive, and accurate comparative modeling pipeline applicable to whole genomes. Such large-scale modeling is likely to encourage new kinds of applications for the many resulting models, based on their large number and completeness at the level of the family, organism, or functional network. [ABSTRACT FROM AUTHOR]