Selecting targets for structural determination by navigating in a graph of protein families

MOTIVATION: A major goal in structural genomics is to enrich the catalogue of proteins whose 3D structures are known. In an attempt to address this problem we mapped over 10 000 proteins with solved structures onto a graph of all Swissprot protein sequences (release 36, approximately 73 000 proteins) provided by ProtoMap, with the goal of sorting proteins according to their likelihood of belonging to new superfamilies. We hypothesized that proteins within neighbouring clusters tend to share common structural superfamilies or folds. If true, the likelihood of finding new superfamilies increases in clusters that are distal from other solved structures within the graph. RESULTS: We defined an order relation between unsolved proteins according to their 'distance' from solved structures in the graph, and sorted approximately 48 000 proteins. Our list can be partitioned into three groups: approximately 35 000 proteins sharing a cluster with at least one known structure; approximately 6500 proteins in clusters with no solved structure but with neighbouring clusters containing known structures; and a third group contains the rest of the proteins, approximately 6100 (in 1274 clusters). We tested the quality of the order relation using thousands of recently solved structures that were not included when the order was defined. The tests show that our order is significantly better (P-value approximately 10(5)) than a random order. More interestingly, the order within the union of the second and third groups, and the order within the third group alone, perform better than random (P-values: 0.0008 and 0.15, respectively) and are better than alternative orders created using PSI-BLAST. Herein, we present a method for selecting targets to be used in structural genomics projects. AVAILABILITY: List of proteins to be used for targets selection combined with a set of biological filters for narrowing down potential targets is in http://www.protarget.cs.huji.ac.il.     

ProtoNet: hierarchical classification of the protein space

The ProtoNet site provides an automatic hierarchical clustering of the SWISS-PROT protein database. The clustering is based on an all-against-all BLAST similarity search. The similarities' E-score is used to perform a continuous bottom-up clustering process by applying alternative rules for merging clusters. The outcome of this clustering process is a classification of the input proteins into a hierarchy of clusters of varying degrees of granularity. ProtoNet (version 1.3) is accessible in the form of an interactive web site at http://www.protonet.cs.huji.ac.il. ProtoNet provides navigation tools for monitoring the clustering process with a vertical and horizontal view. Each cluster at any level of the hierarchy is assigned with a statistical index, indicating the level of purity based on biological keywords such as those provided by SWISS-PROT and InterPro. ProtoNet can be used for function prediction, for defining superfamilies and subfamilies and for large-scale protein annotation purposes.     

A robust method to detect structural and functional remote homologues

With currently available sequence data, it is feasible to conduct extensive comparisons among large sets of protein sequences. It is still a much more challenging task to partition the protein space into structurally and functionally related families solely based on sequence comparisons. The ProtoNet system automatically generates a treelike classification of the whole protein space. It stands to reason that this classification reflects evolutionary relationships, both close and remote. In this article, we examine this hypothesis. We present a semiautomatic procedure that singles out certain inner nodes in the ProtoNet tree that should ideally correspond to structurally and functionally defined protein families. We compare the performance of this method against several expert systems. Some of the competing methods incorporate additional extraneous information on protein structure or on enzymatic activities. The ProtoNet-based method performs at least as well as any of the methods with which it was compared. This article illustrates the ProtoNet-based method on several evolutionarily diverse families. Using this new method, an evolutionary divergence scheme can be proposed for a large number of structural and functional related superfamilies.     

Fishing with (Proto)Net-a principled approach to protein target selection

Structural genomics strives to represent the entire protein space. The first step towards achieving this goal is by rationally selecting proteins whose structures have not been determined, but that represent an as yet unknown structural superfamily or fold. Once such a structure is solved, it can be used as a template for modelling homologous proteins. This will aid in unveiling the structural diversity of the protein space. Currently, no reliable method for accurate 3D structural prediction is available when a sequence or a structure homologue is not available. Here we present a systematic methodology for selecting target proteins whose structure is likely to adopt a new, as yet unknown superfamily or fold. Our method takes advantage of a global classification of the sequence space as presented by ProtoNet-3D, which is a hierarchical agglomerative clustering of the proteins of interest (the proteins in Swiss-Prot) along with all solved structures (taken from the PDB). By navigating in the scaffold of ProtoNet-3D, we yield a prioritized list of proteins that are not yet structurally solved, along with the probability of each of the proteins belonging to a new superfamily or fold. The sorted list has been self-validated against real structural data that was not available when the predictions were made. The practical application of using our computational-statistical method to determine novel superfamilies for structural genomics projects is also discussed.     

Functional site plasticity in domain superfamilies

We present, to our knowledge, the first quantitative analysis of functional site diversity in homologous domain superfamilies. Different types of functional sites are considered separately. Our results show that most diverse superfamilies are very plastic in terms of the spatial location of their functional sites. This is especially true for protein–protein interfaces. In contrast, we confirm that catalytic sites typically occupy only a very small number of topological locations. Small-ligand binding sites are more diverse than expected, although in a more limited manner than protein–protein interfaces. In spite of the observed diversity, our results also confirm the previously reported preferential location of functional sites. We identify a subset of homologous domain superfamilies where diversity is particularly extreme, and discuss possible reasons for such plasticity, i.e. structural diversity. Our results do not contradict previous reports of preferential co-location of sites among homologues, but rather point at the importance of not ignoring other sites, especially in large and diverse superfamilies. Data on sites exploited by different relatives, within each well annotated domain superfamily, has been made accessible from the CATH website in order to highlight versatile superfamilies or superfamilies with highly preferential sites. This information is valuable for system biology and knowledge of any constraints on protein interactions could help in understanding the dynamic control of networks in which these proteins participate. The novelty of our work lies in the comprehensive nature of the analysis – we have used a significantly larger dataset than previous studies – and the fact that in many superfamilies we show that different parts of the domain surface are exploited by different relatives for ligand/protein interactions, particularly in superfamilies which are diverse in sequence and structure, an observation not previously reported on such a large scale. This article is part of a Special Issue entitled: The emerging dynamic view of proteins: Protein plasticity in allostery, evolution and self-assembly.     

Inference of functional properties from large-scale analysis of enzyme superfamilies

As increasingly large amounts of data from genome and other sequencing projects become available, new approaches are needed to determine the functions of the proteins these genes encode. We show how large-scale computational analysis can help to address this challenge by linking functional information to sequence and structural similarities using protein similarity networks. Network analyses using three functionally diverse enzyme superfamilies illustrate the use of these approaches for facile updating and comparison of available structures for a large superfamily, for creation of functional hypotheses for metagenomic sequences, and to summarize the limits of our functional knowledge about even well studied superfamilies.     

Exploring dynamics of protein structure determination and homology-based prediction to estimate the number of superfamilies and folds

Background

As tertiary structure is currently available only for a fraction of known protein families, it is important to assess what parts of sequence space have been structurally characterized. We consider protein domains whose structure can be predicted by sequence similarity to proteins with solved structure and address the following questions. Do these domains represent an unbiased random sample of all sequence families? Do targets solved by structural genomic initiatives (SGI) provide such a sample? What are approximate total numbers of structure-based superfamilies and folds among soluble globular domains?

Results

To make these assessments, we combine two approaches: (i) sequence analysis and homology-based structure prediction for proteins from complete genomes; and (ii) monitoring dynamics of the assigned structure set in time, with the accumulation of experimentally solved structures. In the Clusters of Orthologous Groups (COG) database, we map the growing population of structurally characterized domain families onto the network of sequence-based connections between domains. This mapping reveals a systematic bias suggesting that target families for structure determination tend to be located in highly populated areas of sequence space. In contrast, the subset of domains whose structure is initially inferred by SGI is similar to a random sample from the whole population. To accommodate for the observed bias, we propose a new non-parametric approach to the estimation of the total numbers of structural superfamilies and folds, which does not rely on a specific model of the sampling process. Based on dynamics of robust distribution-based parameters in the growing set of structure predictions, we estimate the total numbers of superfamilies and folds among soluble globular proteins in the COG database.

Conclusion

The set of currently solved protein structures allows for structure prediction in approximately a third of sequence-based domain families. The choice of targets for structure determination is biased towards domains with many sequence-based homologs. The growing SGI output in the future should further contribute to the reduction of this bias. The total number of structural superfamilies and folds in the COG database are estimated as ~4000 and ~1700. These numbers are respectively four and three times higher than the numbers of superfamilies and folds that can currently be assigned to COG proteins.     

A novel scoring function for predicting the conformations of tightly packed pairs of transmembrane alpha-helices

Pairs of helices in transmembrane (TM) proteins are often tightly packed. We present a scoring function and a computational methodology for predicting the tertiary fold of a pair of alpha-helices such that its chances of being tightly packed are maximized. Since the number of TM protein structures solved to date is small, it seems unlikely that a reliable scoring function derived statistically from the known set of TM protein structures will be available in the near future. We therefore constructed a scoring function based on the qualitative insights gained in the past two decades from the solved structures of TM and soluble proteins. In brief, we reward the formation of contacts between small amino acid residues such as Gly, Cys, and Ser, that are known to promote dimerization of helices, and penalize the burial of large amino acid residues such as Arg and Trp. As a case study, we show that our method predicts the native structure of the TM homodimer glycophorin A (GpA) to be, in essence, at the global score optimum. In addition, by correlating our results with empirical point mutations on this homodimer, we demonstrate that our method can be a helpful adjunct to mutation analysis. We present a data set of canonical alpha-helices from the solved structures of TM proteins and provide a set of programs for analyzing it (http://ashtoret.tau.ac.il/~sarel). From this data set we derived 11 helix pairs, and conducted searches around their native states as a further test of our method. Approximately 73% of our predictions showed a reasonable fit (RMS deviation <2A) with the native structures compared to the success rate of 8% expected by chance. The search method we employ is less effective for helix pairs that are connected via short loops (<20 amino acid residues), indicating that short loops may play an important role in determining the conformation of alpha-helices in TM proteins.     

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.     

Efficient prediction of nucleic acid binding function from low-resolution protein structures

Structural genomics projects as well as ab initio protein structure prediction methods provide structures of proteins with no sequence or fold similarity to proteins with known functions. These are often low-resolution structures that may only include the positions of C alpha atoms. We present a fast and efficient method to predict DNA-binding proteins from just the amino acid sequences and low-resolution, C alpha-only protein models. The method uses the relative proportions of certain amino acids in the protein sequence, the asymmetry of the spatial distribution of certain other amino acids as well as the dipole moment of the molecule. These quantities are used in a linear formula, with coefficients derived from logistic regression performed on a training set, and DNA-binding is predicted based on whether the result is above a certain threshold. We show that the method is insensitive to errors in the atomic coordinates and provides correct predictions even on inaccurate protein models. We demonstrate that the method is capable of predicting proteins with novel binding site motifs and structures solved in an unbound state. The accuracy of our method is close to another, published method that uses all-atom structures, time-consuming calculations and information on conserved residues.     

Predicting Protein Function and Binding Profile via Matching of Local Evolutionary and Geometric Surface Patterns

Inferring protein functions from structures is a challenging task, as a large number of orphan protein structures from structural genomics project are now solved without their biochemical functions characterized. For proteins binding to similar substrates or ligands and carrying out similar functions, their binding surfaces are under similar physicochemical constraints, and hence the sets of allowed and forbidden residue substitutions are similar. However, it is difficult to isolate such selection pressure due to protein function from selection pressure due to protein folding, and evolutionary relationship reflected by global sequence and structure similarities between proteins is often unreliable for inferring protein function. We have developed a method, called pevoSOAR (pocket-based evolutionary search of amino acid residues), for predicting protein functions by solving the problem of uncovering amino acids residue substitution pattern due to protein function and separating it from amino acids substitution pattern due to protein folding. We incorporate evolutionary information specific to an individual binding region and match local surfaces on a large scale with millions of precomputed protein surfaces to identify those with similar functions. Our pevoSOAR method also generates a probablistic model called the computed binding a profile that characterizes protein-binding activities that may involve multiple substrates or ligands. We show that our method can be used to predict enzyme functions with accuracy. Our method can also assess enzyme binding specificity and promiscuity. In an objective large-scale test of 100 enzyme families with thousands of structures, our predictions are found to be sensitive and specific: At the stringent specificity level of 99.98%, we can correctly predict enzyme functions for 80.55% of the proteins. The overall area under the receiver operating characteristic curve measuring the performance of our prediction is 0.955, close to the perfect value of 1.00. The best Matthews coefficient is 86.6%. Our method also works well in predicting the biochemical functions of orphan proteins from structural genomics projects.     

Conformational characterization of disulfide bonds: A tool for protein classification

Background

Throughout evolution, mutations in particular regions of some protein structures have resulted in extra covalent bonds that increase the overall robustness of the fold: disulfide bonds. The two strategically placed cysteines can also have a more direct role in protein function, either by assisting thiol or disulfide exchange, or through allosteric effects. In this work, we verified how the structural similarities between disulfides can reflect functional and evolutionary relationships between different proteins. We analyzed the conformational patterns of the disulfide bonds in a set of disulfide-rich proteins that included twelve SCOP superfamilies: thioredoxin-like and eleven superfamilies containing small disulfide-rich proteins (SDP).

Results

The twenty conformations considered in the present study were characterized by both structural and energetic parameters. The corresponding frequencies present diverse patterns for the different superfamilies. The least-strained conformations are more abundant for the SDP superfamilies, while the “catalytic” +/−RHook is dominant for the thioredoxin-like superfamily. The “allosteric” -RHSaple is moderately abundant for BBI, Crisp and Thioredoxin-like superfamilies and less frequent for the remaining superfamilies. Using a hierarchical clustering analysis we found that the twelve superfamilies were grouped in biologically significant clusters.

Conclusions

In this work, we carried out an extensive statistical analysis of the conformational motifs for the disulfide bonds present in a set of disulfide-rich proteins. We show that the conformational patterns observed in disulfide bonds are sufficient to group proteins that share both functional and structural patterns and can therefore be used as a criterion for protein classification.     

Folding studies of immunoglobulin-like beta-sandwich proteins suggest that they share a common folding pathway.

BACKGROUND: Are folding pathways conserved in protein families? To test this explicitly and ask to what extent structure specifies folding pathways requires comparison of proteins with a common fold. Our strategy is to choose members of a highly diverse protein family with no conservation of function and little or no sequence identity, but with structures that are essentially the same. The immunoglobulin-like fold is one of the most common structural families, and is subdivided into superfamilies with no detectable evolutionary or functional relationship. RESULTS: We compared the folding of a number of immunoglobulin-like proteins that have a common structural core and found a strong correlation between folding rate and stability. The results suggest that the folding pathways of these immunoglobulin-like proteins share common features. CONCLUSIONS: This study is the first to compare the folding of structurally related proteins that are members of different superfamilies. The most likely explanation for the results is that interactions that are important in defining the structure of immunoglobulin-like proteins are also used to guide folding.     

Functional classification of protein structures by local structure matching in graph representation

As a result of high‐throughput protein structure initiatives, over 14,400 protein structures have been solved by Structural Genomics (SG) centers and participating research groups. While the totality of SG data represents a tremendous contribution to genomics and structural biology, reliable functional information for these proteins is generally lacking. Better functional predictions for SG proteins will add substantial value to the structural information already obtained. Our method described herein, Graph Representation of Active Sites for Prediction of Function (GRASP‐Func), predicts quickly and accurately the biochemical function of proteins by representing residues at the predicted local active site as graphs rather than in Cartesian coordinates. We compare the GRASP‐Func method to our previously reported method, Structurally Aligned Local Sites of Activity (SALSA), using the Ribulose Phosphate Binding Barrel (RPBB), 6‐Hairpin Glycosidase (6‐HG), and Concanavalin A‐like Lectins/Glucanase (CAL/G) superfamilies as test cases. In each of the superfamilies, SALSA and the much faster method GRASP‐Func yield similar correct classification of previously characterized proteins, providing a validated benchmark for the new method. In addition, we analyzed SG proteins using our SALSA and GRASP‐Func methods to predict function. Forty‐one SG proteins in the RPBB superfamily, nine SG proteins in the 6‐HG superfamily, and one SG protein in the CAL/G superfamily were successfully classified into one of the functional families in their respective superfamily by both methods. This improved, faster, validated computational method can yield more reliable predictions of function that can be used for a wide variety of applications by the community.     

A functional hierarchical organization of the protein sequence space

Background

It is a major challenge of computational biology to provide a comprehensive functional classification of all known proteins. Most existing methods seek recurrent patterns in known proteins based on manually-validated alignments of known protein families. Such methods can achieve high sensitivity, but are limited by the necessary manual labor. This makes our current view of the protein world incomplete and biased. This paper concerns ProtoNet, a automatic unsupervised global clustering system that generates a hierarchical tree of over 1,000,000 proteins, based solely on sequence similarity.

Results

In this paper we show that ProtoNet correctly captures functional and structural aspects of the protein world. Furthermore, a novel feature is an automatic procedure that reduces the tree to 12% its original size. This procedure utilizes only parameters intrinsic to the clustering process. Despite the substantial reduction in size, the system's predictive power concerning biological functions is hardly affected. We then carry out an automatic comparison with existing functional protein annotations. Consequently, 78% of the clusters in the compressed tree (5,300 clusters) get assigned a biological function with a high confidence. The clustering and compression processes are unsupervised, and robust.

Conclusions

We present an automatically generated unbiased method that provides a hierarchical classification of all currently known proteins.

     

Computer-based screening of functional conformers of proteins

A long-standing goal in biology is to establish the link between function, structure, and dynamics of proteins. Considering that protein function at the molecular level is understood by the ability of proteins to bind to other molecules, the limited structural data of proteins in association with other bio-molecules represents a major hurdle to understanding protein function at the structural level. Recent reports show that protein function can be linked to protein structure and dynamics through network centrality analysis, suggesting that the structures of proteins bound to natural ligands may be inferred computationally. In the present work, a new method is described to discriminate protein conformations relevant to the specific recognition of a ligand. The method relies on a scoring system that matches critical residues with central residues in different structures of a given protein. Central residues are the most traversed residues with the same frequency in networks derived from protein structures. We tested our method in a set of 24 different proteins and more than 260,000 structures of these in the absence of a ligand or bound to it. To illustrate the usefulness of our method in the study of the structure/dynamics/function relationship of proteins, we analyzed mutants of the yeast TATA-binding protein with impaired DNA binding. Our results indicate that critical residues for an interaction are preferentially found as central residues of protein structures in complex with a ligand. Thus, our scoring system effectively distinguishes protein conformations relevant to the function of interest.     

SMpred: a support vector machine approach to identify structural motifs in protein structure without using evolutionary information

Knowledge of three dimensional structure is essential to understand the function of a protein. Although the overall fold is made from the whole details of its sequence, a small group of residues, often called as structural motifs, play a crucial role in determining the protein fold and its stability. Identification of such structural motifs requires sufficient number of sequence and structural homologs to define conservation and evolutionary information. Unfortunately, there are many structures in the protein structure databases have no homologous structures or sequences. In this work, we report an SVM method, SMpred, to identify structural motifs from single protein structure without using sequence and structural homologs. SMpred method was trained and tested using 132 proteins domains containing 581 motifs. SMpred method achieved 78.79% accuracy with 79.06% sensitivity and 78.53% specificity. The performance of SMpred was evaluated with MegaMotifBase using 188 proteins containing 1161 motifs. Out of 1161 motifs, SMpred correctly identified 1503 structural motifs reported in MegaMotifBase. Further, we showed that SMpred is useful approach for the length deviant superfamilies and single member superfamilies. This result suggests the usefulness of our approach for facilitating the identification of structural motifs in protein structure in the absence of sequence and structural homologs. The dataset and executable for the SMpred algorithm is available at http://www3.ntu.edu.sg/home/EPNSugan/index_files/SMpred.htm.     

PASS2: a semi-automated database of Protein Alignments Organised as Structural Superfamilies

PASS2 is a nearly automated version of CAMPASS and contains sequence alignments of proteins grouped at the level of superfamilies. This database has been created to fall in correspondence with SCOP database (1.53 release) and currently consists of 110 multi-member superfamilies and 613 superfamilies corresponding to single members. In multi-member superfamilies, protein chains with no more than 25% sequence identity have been considered for the alignment and hence the database aims to address sequence alignments which represent 26 219 protein domains under the SCOP 1.53 release. Structure-based sequence alignments have been obtained by COMPARER and the initial equivalences are provided automatically from a MALIGN alignment and subsequently augmented using STAMP4.0. The final sequence alignments have been annotated for the structural features using JOY4.0. Several interesting links are provided to other related databases and genome sequence relatives. Availability of reliable sequence alignments of distantly related proteins, despite poor sequence identity and single-member superfamilies, permit better sampling of structures in libraries for fold recognition of new sequences and for the understanding of protein structure–function relationships of individual superfamilies. The database can be queried by keywords and also by sequence search, interfaced by PSI-BLAST methods. Structure-annotated sequence alignments and several structural accessory files can be retrieved for all the superfamilies including the user-input sequence. The database can be accessed from http://www.ncbs.res.in/%7Efaculty/mini/campass/pass.html.     

Identification of homology in protein structure classification.

Structural biology and structural genomics are expected to produce many three-dimensional protein structures in the near future. Each new structure raises questions about its function and evolution. Correct functional and evolutionary classification of a new structure is difficult for distantly related proteins and error-prone using simple statistical scores based on sequence or structure similarity. Here we present an accurate numerical method for the identification of evolutionary relationships (homology). The method is based on the principle that natural selection maintains structural and functional continuity within a diverging protein family. The problem of different rates of structural divergence between different families is solved by first using structural similarities to produce a global map of folds in protein space and then further subdividing fold neighborhoods into superfamilies based on functional similarities. In a validation test against a classification by human experts (SCOP), 77% of homologous pairs were identified with 92% reliability. The method is fully automated, allowing fast, self-consistent and complete classification of large numbers of protein structures. In particular, the discrimination between analogy and homology of close structural neighbors will lead to functional predictions while avoiding overprediction.     

FuzzyART neural network for protein classification

One of the major research directions in bioinformatics is that of predicting the protein superfamily in large databases and classifying a given set of protein domains into superfamilies. The classification reflects the structural, evolutionary and functional relatedness. These relationships are embodied in hierarchical classification such as Structural Classification of Protein (SCOP), which is manually curated. Such classification is essential for the structural and functional analysis of proteins. Yet, a large number of proteins remain unclassified. We have proposed an unsupervised machine-learning FuzzyART neural network algorithm to classify a given set of proteins into SCOP superfamilies. The proposed method is fast learning and uses an atypical non-linear pattern recognition technique. In this approach, we have constructed a similarity matrix from p-values of BLAST all-against-all, trained the network with FuzzyART unsupervised learning algorithm using the similarity matrix as input vectors and finally the trained network offers SCOP superfamily level classification. In this experiment, we have evaluated the performance of our method with existing techniques on six different datasets. We have shown that the trained network is able to classify a given similarity matrix of a set of sequences into SCOP superfamilies at high classification accuracy.