FireDB--a database of functionally important residues from proteins of known structure

GRaSP-web: a machine learning strategy to predict binding sites based on residue neighborhood graphs

Proteins are essential macromolecules for the maintenance of living systems. Many of them perform their function by interacting with other molecules in regions called binding sites. The identification and characterization of these regions are of fundamental importance to determine protein function, being a fundamental step in processes such as drug design and discovery. However, identifying such binding regions is not trivial due to the drawbacks of experimental methods, which are costly and time-consuming. Here we propose GRaSP-web, a web server that uses GRaSP (Graph-based Residue neighborhood Strategy to Predict binding sites), a residue-centric method based on graphs that uses machine learning to predict putative ligand binding site residues. The method outperformed 6 state-of-the-art residue-centric methods (MCC of 0.61). Also, GRaSP-web is scalable as it takes 10-20 seconds to predict binding sites for a protein complex (the state-of-the-art residue-centric method takes 2-5h on the average). It proved to be consistent in predicting binding sites for bound/unbound structures (MCC 0.61 for both) and for a large dataset of multi-chain proteins (4500 entries, MCC 0.61). GRaSPWeb is freely available at https://grasp.ufv.br.

Proteins and Their Interacting Partners: An Introduction to Protein-Ligand Binding Site Prediction Methods with a Focus on FunFOLD3

Proteins are essential molecules with a diverse range of functions; elucidating their biological and biochemical characteristics can be difficult and time consuming using in vitro and/or in vivo methods. Additionally, in vivo protein-ligand binding site elucidation is unable to keep place with current growth in sequencing, leaving the majority of new protein sequences without known functions. Therefore, the development of new methods, which aim to predict the protein-ligand interactions and ligand-binding site residues directly from amino acid sequences, is becoming increasingly important. In silico prediction can utilise either sequence information, structural information or a combination of both. In this chapter, we will discuss the broad range of methods for ligand-binding site prediction from protein structure and we will describe our method, FunFOLD3, for the prediction of protein-ligand interactions and ligand-binding sites based on template-based modelling. Additionally, we will describe the step-by-step instructions using the FunFOLD3 downloadable application along with examples from the Critical Assessment of Techniques for Protein Structure Prediction (CASP) where FunFOLD3 has been used to aid ligand and ligand-binding site prediction. Finally, we will introduce our newer method, FunFOLD3-D, a version of FunFOLD3 which aims to improve template-based protein-ligand binding site prediction through the integration of docking, using AutoDock Vina.

GRaSP: a graph-based residue neighborhood strategy to predict binding sites

Motivation

The discovery of protein–ligand-binding sites is a major step for elucidating protein function and for investigating new functional roles. Detecting protein–ligand-binding sites experimentally is time-consuming and expensive. Thus, a variety of in silico methods to detect and predict binding sites was proposed as they can be scalable, fast and present low cost.

Results

We proposed Graph-based Residue neighborhood Strategy to Predict binding sites (GRaSP), a novel residue centric and scalable method to predict ligand-binding site residues. It is based on a supervised learning strategy that models the residue environment as a graph at the atomic level. Results show that GRaSP made compatible or superior predictions when compared with methods described in the literature. GRaSP outperformed six other residue-centric methods, including the one considered as state-of-the-art. Also, our method achieved better results than the method from CAMEO independent assessment. GRaSP ranked second when compared with five state-of-the-art pocket-centric methods, which we consider a significant result, as it was not devised to predict pockets. Finally, our method proved scalable as it took 10–20 s on average to predict the binding site for a protein complex whereas the state-of-the-art residue-centric method takes 2–5 h on average.

Availability and implementation

The source code and datasets are available at https://github.com/charles-abreu/GRaSP.

Supplementary information

Supplementary data are available at Bioinformatics online.

CavBench: A benchmark for protein cavity detection methods

Extensive research has been applied to discover new techniques and methods to model protein-ligand interactions. In particular, considerable efforts focused on identifying candidate binding sites, which quite often are active sites that correspond to protein pockets or cavities. Thus, these cavities play an important role in molecular docking. However, there is no established benchmark to assess the accuracy of new cavity detection methods. In practice, each new technique is evaluated using a small set of proteins with known binding sites as ground-truth. However, studies supported by large datasets of known cavities and/or binding sites and statistical classification (i.e., false positives, false negatives, true positives, and true negatives) would yield much stronger and reliable assessments. To this end, we propose CavBench, a generic and extensible benchmark to compare different cavity detection methods relative to diverse ground truth datasets (e.g., PDBsum) using statistical classification methods.

Effect of the sequence data deluge on the performance of methods for detecting protein functional residues

BACKGROUND

The exponential accumulation of new sequences in public databases is expected to improve the performance of all the approaches for predicting protein structural and functional features. Nevertheless, this was never assessed or quantified for some widely used methodologies, such as those aimed at detecting functional sites and functional subfamilies in protein multiple sequence alignments. Using raw protein sequences as only input, these approaches can detect fully conserved positions, as well as those with a family-dependent conservation pattern. Both types of residues are routinely used as predictors of functional sites and, consequently, understanding how the sequence content of the databases affects them is relevant and timely.

RESULTS

In this work we evaluate how the growth and change with time in the content of sequence databases affect five sequence-based approaches for detecting functional sites and subfamilies. We do that by recreating historical versions of the multiple sequence alignments that would have been obtained in the past based on the database contents at different time points, covering a period of 20 years. Applying the methods to these historical alignments allows quantifying the temporal variation in their performance. Our results show that the number of families to which these methods can be applied sharply increases with time, while their ability to detect potentially functional residues remains almost constant.

CONCLUSIONS

These results are informative for the methods' developers and final users, and may have implications in the design of new sequencing initiatives.

Identification of Protein-Ligand Binding Sites by Sequence Information and Ensemble Classifier

Identifying protein-ligand binding sites is an important process in drug discovery and structure-based drug design. Detecting protein-ligand binding sites is expensive and time-consuming by traditional experimental methods. Hence, computational approaches provide many effective strategies to deal with this issue. Recently, lots of computational methods are based on structure information on proteins. However, these methods are limited in the common scenario, where both the sequence of protein target is known and sufficient 3D structure information is available. Studies indicate that sequence-based computational approaches for predicting protein-ligand binding sites are more practical. In this paper, we employ a novel computational model of protein-ligand binding sites prediction, using protein sequence. We apply the Discrete Cosine Transform (DCT) to extract feature from Position-Specific Score Matrix (PSSM). In order to improve the accuracy, Predicted Relative Solvent Accessibility (PRSA) information is also utilized. The predictor of protein-ligand binding sites is built by employing the ensemble weighted sparse representation model with random under-sampling. To evaluate our method, we conduct several comprehensive tests (12 types of ligands testing sets) for predicting protein-ligand binding sites. Results show that our method achieves better Matthew's correlation coefficient (MCC) than other outstanding methods on independent test sets of ATP (0.506), ADP (0.511), AMP (0.393), GDP (0.579), GTP (0.641), Mg²⁺ (0.317), Fe³⁺ (0.490) and HEME (0.640). Our proposed method outperforms earlier predictors (the performance of MCC) in 8 of the 12 ligands types.

KinMutRF: a random forest classifier of sequence variants in the human protein kinase superfamily

Background

The association between aberrant signal processing by protein kinases and human diseases such as cancer was established long time ago. However, understanding the link between sequence variants in the protein kinase superfamily and the mechanistic complex traits at the molecular level remains challenging: cells tolerate most genomic alterations and only a minor fraction disrupt molecular function sufficiently and drive disease.

Results

KinMutRF is a novel random-forest method to automatically identify pathogenic variants in human kinases. Twenty six decision trees implemented as a random forest ponder a battery of features that characterize the variants: a) at the gene level, including membership to a Kinbase group and Gene Ontology terms; b) at the PFAM domain level; and c) at the residue level, the types of amino acids involved, changes in biochemical properties, functional annotations from UniProt, Phospho.ELM and FireDB. KinMutRF identifies disease-associated variants satisfactorily (Acc: 0.88, Prec:0.82, Rec:0.75, F-score:0.78, MCC:0.68) when trained and cross-validated with the 3689 human kinase variants from UniProt that have been annotated as neutral or pathogenic. All unclassified variants were excluded from the training set. Furthermore, KinMutRF is discussed with respect to two independent kinase-specific sets of mutations no included in the training and testing, Kin-Driver (643 variants) and Pon-BTK (1495 variants). Moreover, we provide predictions for the 848 protein kinase variants in UniProt that remained unclassified.

A public implementation of KinMutRF, including documentation and examples, is available online (http://kinmut2.bioinfo.cnio.es). The source code for local installation is released under a GPL version 3 license, and can be downloaded from https://github.com/Rbbt-Workflows/KinMut2.

Conclusions

KinMutRF is capable of classifying kinase variation with good performance. Predictions by KinMutRF compare favorably in a benchmark with other state-of-the-art methods (i.e. SIFT, Polyphen-2, MutationAssesor, MutationTaster, LRT, CADD, FATHMM, and VEST). Kinase-specific features rank as the most elucidatory in terms of information gain and are likely the improvement in prediction performance. This advocates for the development of family-specific classifiers able to exploit the discriminatory power of features unique to individual protein families.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-016-2723-1) contains supplementary material, which is available to authorized users.

Impact of germline and somatic missense variations on drug binding sites

Advancements in next-generation sequencing (NGS) technologies are generating a vast amount of data. This exacerbates the current challenge of translating NGS data into actionable clinical interpretations. We have comprehensively combined germline and somatic nonsynonymous single-nucleotide variations (nsSNVs) that affect drug binding sites in order to investigate their prevalence. The integrated data thus generated in conjunction with exome or whole-genome sequencing can be used to identify patients who may not respond to a specific drug because of alterations in drug binding efficacy due to nsSNVs in the target protein's gene. To identify the nsSNVs that may affect drug binding, protein–drug complex structures were retrieved from Protein Data Bank (PDB) followed by identification of amino acids in the protein–drug binding sites using an occluded surface method. Then, the germline and somatic mutations were mapped to these amino acids to identify which of these alter protein–drug binding sites. Using this method we identified 12 993 amino acid–drug binding sites across 253 unique proteins bound to 235 unique drugs. The integration of amino acid–drug binding sites data with both germline and somatic nsSNVs data sets revealed 3133 nsSNVs affecting amino acid–drug binding sites. In addition, a comprehensive drug target discovery was conducted based on protein structure similarity and conservation of amino acid–drug binding sites. Using this method, 81 paralogs were identified that could serve as alternative drug targets. In addition, non-human mammalian proteins bound to drugs were used to identify 142 homologs in humans that can potentially bind to drugs. In the current protein–drug pairs that contain somatic mutations within their binding site, we identified 85 proteins with significant differential gene expression changes associated with specific cancer types. Information on protein–drug binding predicted drug target proteins and prevalence of both somatic and germline nsSNVs that disrupt these binding sites can provide valuable knowledge for personalized medicine treatment. A web portal is available where nsSNVs from individual patient can be checked by scanning against DrugVar to determine whether any of the SNVs affect the binding of any drug in the database.

Computer analysis of protein functional sites projection on exon structure of genes in Metazoa

BACKGROUND

Study of the relationship between the structural and functional organization of proteins and their coding genes is necessary for an understanding of the evolution of molecular systems and can provide new knowledge for many applications for designing proteins with improved medical and biological properties. It is well known that the functional properties of proteins are determined by their functional sites. Functional sites are usually represented by a small number of amino acid residues that are distantly located from each other in the amino acid sequence. They are highly conserved within their functional group and vary significantly in structure between such groups. According to this facts analysis of the general properties of the structural organization of the functional sites at the protein level and, at the level of exon-intron structure of the coding gene is still an actual problem.

RESULTS

One approach to this analysis is the projection of amino acid residue positions of the functional sites along with the exon boundaries to the gene structure. In this paper, we examined the discontinuity of the functional sites in the exon-intron structure of genes and the distribution of lengths and phases of the functional site encoding exons in vertebrate genes. We have shown that the DNA fragments coding the functional sites were in the same exons, or in close exons. The observed tendency to cluster the exons that code functional sites which could be considered as the unit of protein evolution. We studied the characteristics of the structure of the exon boundaries that code, and do not code, functional sites in 11 Metazoa species. This is accompanied by a reduced frequency of intercodon gaps (phase 0) in exons encoding the amino acid residue functional site, which may be evidence of the existence of evolutionary limitations to the exon shuffling.

CONCLUSIONS

These results characterize the features of the coding exon-intron structure that affect the functionality of the encoded protein and allow a better understanding of the emergence of biological diversity.

Constructing query-driven dynamic machine learning model with application to protein-ligand binding sites prediction

We are facing an era with annotated biological data rapidly and continuously generated. How to effectively incorporate new annotated data into the learning step is crucial for enhancing the performance of a bioinformatics prediction model. Although machine-learning-based methods have been extensively used for dealing with various biological problems, existing approaches usually train static prediction models based on fixed training datasets. The static approaches are found having several disadvantages such as low scalability and impractical when training dataset is huge. In view of this, we propose a dynamic learning framework for constructing query-driven prediction models. The key difference between the proposed framework and the existing approaches is that the training set for the machine learning algorithm of the proposed framework is dynamically generated according to the query input, as opposed to training a general model regardless of queries in traditional static methods. Accordingly, a query-driven predictor based on the smaller set of data specifically selected from the entire annotated base dataset will be applied on the query. The new way for constructing the dynamic model enables us capable of updating the annotated base dataset flexibly and using the most relevant core subset as the training set makes the constructed model having better generalization ability on the query, showing "part could be better than all" phenomenon. According to the new framework, we have implemented a dynamic protein-ligand binding sites predictor called OSML (On-site model for ligand binding sites prediction). Computer experiments on 10 different ligand types of three hierarchically organized levels show that OSML outperforms most existing predictors. The results indicate that the current dynamic framework is a promising future direction for bridging the gap between the rapidly accumulated annotated biological data and the effective machine-learning-based predictors. OSML web server and datasets are freely available at: http://www.csbio.sjtu.edu.cn/bioinf/OSML/ for academic use.

wKinMut-2: Identification and Interpretation of Pathogenic Variants in Human Protein Kinases

Most genomic alterations are tolerated while only a minor fraction disrupts molecular function sufficiently to drive disease. Protein kinases play a central biological function and the functional consequences of their variants are abundantly characterized. However, this heterogeneous information is often scattered across different sources, which makes the integrative analysis complex and laborious. wKinMut-2 constitutes a solution to facilitate the interpretation of the consequences of human protein kinase variation. Nine methods predict their pathogenicity, including a kinase-specific random forest approach. To understand the biological mechanisms causative of human diseases and cancer, information from pertinent reference knowledge bases and the literature is automatically mined, digested, and homogenized. Variants are visualized in their structural contexts and residues affecting catalytic and drug binding are identified. Known protein-protein interactions are reported. Altogether, this information is intended to assist the generation of new working hypothesis to be corroborated with ulterior experimental work. The wKinMut-2 system, along with a user manual and examples, is freely accessible at http://kinmut2.bioinfo.cnio.es, the code for local installations can be downloaded from https://github.com/Rbbt-Workflows/KinMut2.

A fast topological analysis algorithm for large-scale similarity evaluations of ligands and binding pockets

Motivation

With the rapid increase of the structural data of biomolecular complexes, novel structural analysis methods have to be devised with high-throughput capacity to handle immense data input and to construct massive networks at the minimal computational cost. Moreover, novel methods should be capable of handling a broad range of molecular structural sizes and chemical natures, cognisant of the conformational and electrostatic bases of molecular recognition, and sufficiently accurate to enable contextually relevant biological inferences.

Results

A novel molecular topology comparison method was developed and tested. The method was tested for both ligand and binding pocket similarity analyses and a PDB-wide ligand topological similarity map was computed.

Conclusion

The unprecedentedly wide scope of ligand definition and large-scale topological similarity mapping can provide very robust tools, of performance unmatched by the present alignment-based methods. The method remarkably shows potential for application for scaffold hopping purposes. It also opens new frontiers in the areas of ligand-mediated protein connectivity, ligand-based molecular phylogeny, target fishing, and off-target predictions.

Electronic supplementary material

The online version of this article (doi:10.1186/s13321-015-0091-5) contains supplementary material, which is available to authorized users.

In the multi-domain protein adenylate kinase, domain insertion facilitates cooperative folding while accommodating function at domain interfaces

Having multiple domains in proteins can lead to partial folding and increased aggregation. Folding cooperativity, the all or nothing folding of a protein, can reduce this aggregation propensity. In agreement with bulk experiments, a coarse-grained structure-based model of the three-domain protein, E. coli Adenylate kinase (AKE), folds cooperatively. Domain interfaces have previously been implicated in the cooperative folding of multi-domain proteins. To understand their role in AKE folding, we computationally create mutants with deleted inter-domain interfaces and simulate their folding. We find that inter-domain interfaces play a minor role in the folding cooperativity of AKE. On further analysis, we find that unlike other multi-domain proteins whose folding has been studied, the domains of AKE are not singly-linked. Two of its domains have two linkers to the third one, i.e., they are inserted into the third one. We use circular permutation to modify AKE chain-connectivity and convert inserted-domains into singly-linked domains. We find that domain insertion in AKE achieves the following: (1) It facilitates folding cooperativity even when domains have different stabilities. Insertion constrains the N- and C-termini of inserted domains and stabilizes their folded states. Therefore, domains that perform conformational transitions can be smaller with fewer stabilizing interactions. (2) Inter-domain interactions are not needed to promote folding cooperativity and can be tuned for function. In AKE, these interactions help promote conformational dynamics limited catalysis. Finally, using structural bioinformatics, we suggest that domain insertion may also facilitate the cooperative folding of other multi-domain proteins.

Most individual protein domains fold in an all or nothing fashion. This cooperative folding is important because it reduces the existence of partially folded proteins which can stick to each other and create disease causing aggregates. However, numerous proteins have multiple domains, independent units of folding, stability and/or function. Several such proteins also fold cooperatively. It is thought that strong interactions between individual domains allow the folding to propagate from a nucleating domain to neighbouring ones and this enables cooperative folding in multi-domain proteins. Here, we computationally study the folding of the three-domain protein AKE and find instead that the topology of the protein, wherein the two less stable domains are inserted into the more stable one, promotes folding cooperativity. When the more stable domain is folded, the ends of the inserted domains are constrained and this allows them to fold easily. In such a protein topology, strong inter-domain interactions are not needed to promote folding cooperativity. Interface amino acids which would have been involved in ensuring that the domains fit together correctly can now be tuned for binding or catalysis or conformational transitions. Thus, inserted domains may be present in multi-domain proteins to promote both function and folding.

Assessment of ligand binding site predictions in CASP10

The identification of amino acid residues in proteins involved in binding small molecule ligands is an important step for their functional characterization, as the function of a protein often depends on specific interactions with other molecules. The accuracy of computational methods aiming to predict such binding residues was evaluated within the "function prediction (prediction of binding sites, FN)" category of the critical assessment of protein structure prediction (CASP) experiment. In the last edition of the experiment (CASP10), 17 research groups participated in this category, and their predictions were evaluated on 13 prediction targets containing biologically relevant ligands. The results of this experiment indicate that several methods achieved an overall good performance, showing the usefulness of such methods in predicting ligand binding residues. As in previous years, methods based on a homology transfer approach were dominating. In comparison to CASP9, a larger fraction of the top predictors are automated servers. However, due to the small number of targets and the characteristics of the prediction format, the differences observed among the first ten methods were not statistically significant and it was also not possible to analyze differences in accuracy for different ligand types or overall structure, difficulty. To overcome these limitations and to allow for a more detailed evaluation, in future editions of CASP, methods in the FN category will no longer be evaluated on the "normal" CASP targets, but assessed continuously by CAMEO (continuous automated model evaluation) based on weekly prereleased sequences from the PDB.

wKinMut: an integrated tool for the analysis and interpretation of mutations in human protein kinases

Background

Protein kinases are involved in relevant physiological functions and a broad number of mutations in this superfamily have been reported in the literature to affect protein function and stability. Unfortunately, the exploration of the consequences on the phenotypes of each individual mutation remains a considerable challenge.

Results

The wKinMut web-server offers direct prediction of the potential pathogenicity of the mutations from a number of methods, including our recently developed prediction method based on the combination of information from a range of diverse sources, including physicochemical properties and functional annotations from FireDB and Swissprot and kinase-specific characteristics such as the membership to specific kinase groups, the annotation with disease-associated GO terms or the occurrence of the mutation in PFAM domains, and the relevance of the residues in determining kinase subfamily specificity from S3Det. This predictor yields interesting results that compare favourably with other methods in the field when applied to protein kinases.

Together with the predictions, wKinMut offers a number of integrated services for the analysis of mutations. These include: the classification of the kinase, information about associations of the kinase with other proteins extracted from iHop, the mapping of the mutations onto PDB structures, pathogenicity records from a number of databases and the classification of mutations in large-scale cancer studies. Importantly, wKinMut is connected with the SNP2L system that extracts mentions of mutations directly from the literature, and therefore increases the possibilities of finding interesting functional information associated to the studied mutations.

Conclusions

wKinMut facilitates the exploration of the information available about individual mutations by integrating prediction approaches with the automatic extraction of information from the literature (text mining) and several state-of-the-art databases.

wKinMut has been used during the last year for the analysis of the consequences of mutations in the context of a number of cancer genome projects, including the recent analysis of Chronic Lymphocytic Leukemia cases and is publicly available at http://wkinmut.bioinfo.cnio.es.

Exploring the composition of protein-ligand binding sites on a large scale

The residue composition of a ligand binding site determines the interactions available for diffusion-mediated ligand binding, and understanding general composition of these sites is of great importance if we are to gain insight into the functional diversity of the proteome. Many structure-based drug design methods utilize such heuristic information for improving prediction or characterization of ligand-binding sites in proteins of unknown function. The Binding MOAD database if one of the largest curated sets of protein-ligand complexes, and provides a source of diverse, high-quality data for establishing general trends of residue composition from currently available protein structures. We present an analysis of 3,295 non-redundant proteins with 9,114 non-redundant binding sites to identify residues over-represented in binding regions versus the rest of the protein surface. The Binding MOAD database delineates biologically-relevant “valid” ligands from “invalid” small-molecule ligands bound to the protein. Invalids are present in the crystallization medium and serve no known biological function. Contacts are found to differ between these classes of ligands, indicating that residue composition of biologically relevant binding sites is distinct not only from the rest of the protein surface, but also from surface regions capable of opportunistic binding of non-functional small molecules. To confirm these trends, we perform a rigorous analysis of the variation of residue propensity with respect to the size of the dataset and the content bias inherent in structure sets obtained from a large protein structure database. The optimal size of the dataset for establishing general trends of residue propensities, as well as strategies for assessing the significance of such trends, are suggested for future studies of binding-site composition.

Describing the general structure of protein binding sites is fundamentally important for guiding drug design and better understanding structure-function relationships. Here, we analyze small molecules bound to proteins within our large database, Binding MOAD (Mother of All Databases, pronounced like “mode” as a pun referring to ligand-binding modes). We focus on different contacts across the residues in the binding sites, and we normalize the data relative to the protein's entire surface. A key feature of this study is the use of a “control” where we compare real, functional binding sites to the random contacts seen for crystallographic additives against the protein surface. Controls are required in experimental biology, but they are ill-defined in many computational approaches. This allows us to describe how true binding sites are unique on the protein surface and distinct from random patches that attract common, small molecules.

FireDB: a compendium of biological and pharmacologically relevant ligands

FireDB (http://firedb.bioinfo.cnio.es) is a curated inventory of catalytic and biologically relevant small ligand-binding residues culled from the protein structures in the Protein Data Bank. Here we present the important new additions since the publication of FireDB in 2007. The database now contains an extensive list of manually curated biologically relevant compounds. Biologically relevant compounds are informative because of their role in protein function, but they are only a small fraction of the entire ligand set. For the remaining ligands, the FireDB provides cross-references to the annotations from publicly available biological, chemical and pharmacological compound databases. FireDB now has external references for 95% of contacting small ligands, making FireDB a more complete database and providing the scientific community with easy access to the pharmacological annotations of PDB ligands. In addition to the manual curation of ligands, FireDB also provides insights into the biological relevance of individual binding sites. Here, biological relevance is calculated from the multiple sequence alignments of related binding sites that are generated from all-against-all comparison of each FireDB binding site. The database can be accessed by RESTful web services and is available for download via MySQL.

Designing template-free predictor for targeting protein-ligand binding sites with classifier ensemble and spatial clustering

Accurately identifying the protein-ligand binding sites or pockets is of significant importance for both protein function analysis and drug design. Although much progress has been made, challenges remain, especially when the 3D structures of target proteins are not available or no homology templates can be found in the library, where the template-based methods are hard to be applied. In this paper, we report a new ligand-specific template-free predictor called TargetS for targeting protein-ligand binding sites from primary sequences. TargetS first predicts the binding residues along the sequence with ligand-specific strategy and then further identifies the binding sites from the predicted binding residues through a recursive spatial clustering algorithm. Protein evolutionary information, predicted protein secondary structure, and ligand-specific binding propensities of residues are combined to construct discriminative features; an improved AdaBoost classifier ensemble scheme based on random undersampling is proposed to deal with the serious imbalance problem between positive (binding) and negative (nonbinding) samples. Experimental results demonstrate that TargetS achieves high performances and outperforms many existing predictors. TargetS web server and data sets are freely available at: http://www.csbio.sjtu.edu.cn/bioinf/TargetS/ for academic use.

An update of DIVERGE software for functional divergence analysis of protein family

DIVERGE is a software system for phylogeny-based analyses of protein family evolution and functional divergence. It provides a suite of statistical tools for selection and prioritization of the amino acid sites that are responsible for the functional divergence of a gene family. The synergistic efforts of DIVERGE and other methods have convincingly demonstrated that the pattern of rate change at a particular amino acid site may contain insightful information about the underlying functional divergence following gene duplication. These predicted sites may be used as candidates for further experiments. We are now releasing an updated version of DIVERGE with the following improvements: 1) a feasible approach to examining functional divergence in nearly complete sequences by including deletions and insertions (indels); 2) the calculation of the false discovery rate of functionally diverging sites; 3) estimation of the effective number of functional divergence-related sites that is reliable and insensitive to cutoffs; 4) a statistical test for asymmetric functional divergence; and 5) a new method to infer functional divergence specific to a given duplicate cluster. In addition, we have made efforts to improve software design and produce a well-written software manual for the general user.

APPRIS: annotation of principal and alternative splice isoforms

Here, we present APPRIS (http://appris.bioinfo.cnio.es), a database that houses annotations of human splice isoforms. APPRIS has been designed to provide value to manual annotations of the human genome by adding reliable protein structural and functional data and information from cross-species conservation. The visual representation of the annotations provided by APPRIS for each gene allows annotators and researchers alike to easily identify functional changes brought about by splicing events. In addition to collecting, integrating and analyzing reliable predictions of the effect of splicing events, APPRIS also selects a single reference sequence for each gene, here termed the principal isoform, based on the annotations of structure, function and conservation for each transcript. APPRIS identifies a principal isoform for 85% of the protein-coding genes in the GENCODE 7 release for ENSEMBL. Analysis of the APPRIS data shows that at least 70% of the alternative (non-principal) variants would lose important functional or structural information relative to the principal isoform.

BioLiP: a semi-manually curated database for biologically relevant ligand-protein interactions

BioLiP (http://zhanglab.ccmb.med.umich.edu/BioLiP/) is a semi-manually curated database for biologically relevant ligand–protein interactions. Establishing interactions between protein and biologically relevant ligands is an important step toward understanding the protein functions. Most ligand-binding sites prediction methods use the protein structures from the Protein Data Bank (PDB) as templates. However, not all ligands present in the PDB are biologically relevant, as small molecules are often used as additives for solving the protein structures. To facilitate template-based ligand–protein docking, virtual ligand screening and protein function annotations, we develop a hierarchical procedure for assessing the biological relevance of ligands present in the PDB structures, which involves a four-step biological feature filtering followed by careful manual verifications. This procedure is used for BioLiP construction. Each entry in BioLiP contains annotations on: ligand-binding residues, ligand-binding affinity, catalytic sites, Enzyme Commission numbers, Gene Ontology terms and cross-links to the other databases. In addition, to facilitate the use of BioLiP for function annotation of uncharacterized proteins, a new consensus-based algorithm COACH is developed to predict ligand-binding sites from protein sequence or using 3D structure. The BioLiP database is updated weekly and the current release contains 204 223 entries.

Interpretation of the consequences of mutations in protein kinases: combined use of bioinformatics and text mining

Protein kinases play a crucial role in a plethora of significant physiological functions and a number of mutations in this superfamily have been reported in the literature to disrupt protein structure and/or function. Computational and experimental research aims to discover the mechanistic connection between mutations in protein kinases and disease with the final aim of predicting the consequences of mutations on protein function and the subsequent phenotypic alterations. In this article, we will review the possibilities and limitations of current computational methods for the prediction of the pathogenicity of mutations in the protein kinase superfamily. In particular we will focus on the problem of benchmarking the predictions with independent gold standard datasets. We will propose a pipeline for the curation of mutations automatically extracted from the literature. Since many of these mutations are not included in the databases that are commonly used to train the computational methods to predict the pathogenicity of protein kinase mutations we propose them to build a valuable gold standard dataset in the benchmarking of a number of these predictors. Finally, we will discuss how text mining approaches constitute a powerful tool for the interpretation of the consequences of mutations in the context of disease genome analysis with particular focus on cancer.

Getting personalized cancer genome analysis into the clinic: the challenges in bioinformatics

Progress in genomics has raised expectations in many fields, and particularly in personalized cancer research. The new technologies available make it possible to combine information about potential disease markers, altered function and accessible drug targets, which, coupled with pathological and medical information, will help produce more appropriate clinical decisions. The accessibility of such experimental techniques makes it all the more necessary to improve and adapt computational strategies to the new challenges. This review focuses on the critical issues associated with the standard pipeline, which includes: DNA sequencing analysis; analysis of mutations in coding regions; the study of genome rearrangements; extrapolating information on mutations to the functional and signaling level; and predicting the effects of therapies using mouse tumor models. We describe the possibilities, limitations and future challenges of current bioinformatics strategies for each of these issues. Furthermore, we emphasize the need for the collaboration between the bioinformaticians who implement the software and use the data resources, the computational biologists who develop the analytical methods, and the clinicians, the systems' end users and those ultimately responsible for taking medical decisions. Finally, the different steps in cancer genome analysis are illustrated through examples of applications in cancer genome analysis.

Prioritization of pathogenic mutations in the protein kinase superfamily

BACKGROUND

Most of the many mutations described in human protein kinases are tolerated without significant disruption of the corresponding structures or molecular functions, while some of them have been associated to a variety of human diseases, including cancer. In the last decade, a plethora of computational methods to predict the effect of missense single-nucleotide variants (SNVs) have been developed. Still, current high-throughput sequencing efforts and the concomitant need for massive interpretation of protein sequence variants will demand for more efficient and/or accurate computational methods in the forthcoming years.

RESULTS

We present KinMut, a support vector machine (SVM) approach, to identify pathogenic mutations in the protein kinase superfamily. KinMut relays on a combination of sequence-derived features that describe mutations at different levels: (1) Gene level: membership to a specific group in Kinbase and the annotation with GO terms; (2) Domain level: annotated PFAM domains; and (3) Residue level: physicochemical features of amino acids, specificity determining positions, and functional annotations from SwissProt and FireDB. The system has been trained with the set of 3492 human kinase mutations in UniProt for which experimental validation of their pathogenic or neutral character exists. In addition, we discuss the relative importance of these independent properties and their combination for the development of a kinase-specific predictor. Finally, we compare KinMut with other state-of-the-art prediction methods.

CONCLUSIONS

Family-specific features appear among the most discriminative information sources, which allow us to produce accurate results in a reliable and very simple way with minimal supervision. Our study aims to broaden the knowledge on the mechanisms by which mutations in the human kinome contribute to disease with a particular focus in cancer. The classifier as well as further documentation is available at http://kinmut.bioinfo.cnio.es/.

Bioinformatics and variability in drug response: a protein structural perspective

Marketed drugs frequently perform worse in clinical practice than in the clinical trials on which their approval is based. Many therapeutic compounds are ineffective for a large subpopulation of patients to whom they are prescribed; worse, a significant fraction of patients experience adverse effects more severe than anticipated. The unacceptable risk-benefit profile for many drugs mandates a paradigm shift towards personalized medicine. However, prior to adoption of patient-specific approaches, it is useful to understand the molecular details underlying variable drug response among diverse patient populations. Over the past decade, progress in structural genomics led to an explosion of available three-dimensional structures of drug target proteins while efforts in pharmacogenetics offered insights into polymorphisms correlated with differential therapeutic outcomes. Together these advances provide the opportunity to examine how altered protein structures arising from genetic differences affect protein-drug interactions and, ultimately, drug response. In this review, we first summarize structural characteristics of protein targets and common mechanisms of drug interactions. Next, we describe the impact of coding mutations on protein structures and drug response. Finally, we highlight tools for analysing protein structures and protein-drug interactions and discuss their application for understanding altered drug responses associated with protein structural variants.

The Proteome Folding Project: proteome-scale prediction of structure and function

The incompleteness of proteome structure and function annotation is a critical problem for biologists and, in particular, severely limits interpretation of high-throughput and next-generation experiments. We have developed a proteome annotation pipeline based on structure prediction, where function and structure annotations are generated using an integration of sequence comparison, fold recognition, and grid-computing-enabled de novo structure prediction. We predict protein domain boundaries and three-dimensional (3D) structures for protein domains from 94 genomes (including human, Arabidopsis, rice, mouse, fly, yeast, Escherichia coli, and worm). De novo structure predictions were distributed on a grid of more than 1.5 million CPUs worldwide (World Community Grid). We generated significant numbers of new confident fold annotations (9% of domains that are otherwise unannotated in these genomes). We demonstrate that predicted structures can be combined with annotations from the Gene Ontology database to predict new and more specific molecular functions.

Characterization of pathogenic germline mutations in human protein kinases

BACKGROUND

Protein Kinases are a superfamily of proteins involved in crucial cellular processes such as cell cycle regulation and signal transduction. Accordingly, they play an important role in cancer biology. To contribute to the study of the relation between kinases and disease we compared pathogenic mutations to neutral mutations as an extension to our previous analysis of cancer somatic mutations. First, we analyzed native and mutant proteins in terms of amino acid composition. Secondly, mutations were characterized according to their potential structural effects and finally, we assessed the location of the different classes of polymorphisms with respect to kinase-relevant positions in terms of subfamily specificity, conservation, accessibility and functional sites.

RESULTS

Pathogenic Protein Kinase mutations perturb essential aspects of protein function, including disruption of substrate binding and/or effector recognition at family-specific positions. Interestingly these mutations in Protein Kinases display a tendency to avoid structurally relevant positions, what represents a significant difference with respect to the average distribution of pathogenic mutations in other protein families.

CONCLUSIONS

Disease-associated mutations display sound differences with respect to neutral mutations: several amino acids are specific of each mutation type, different structural properties characterize each class and the distribution of pathogenic mutations within the consensus structure of the Protein Kinase domain is substantially different to that for non-pathogenic mutations. This preferential distribution confirms previous observations about the functional and structural distribution of the controversial cancer driver and passenger somatic mutations and their use as a proxy for the study of the involvement of somatic mutations in cancer development.

firestar--advances in the prediction of functionally important residues

firestar is a server for predicting catalytic and ligand-binding residues in protein sequences. Here, we present the important developments since the first release of firestar. Previous versions of the server required human interpretation of the results; the server is now fully automatized. firestar has been implemented as a web service and can now be run in high-throughput mode. Prediction coverage has been greatly improved with the extension of the FireDB database and the addition of alignments generated by HHsearch. Ligands in FireDB are now classified for biological relevance. Many of the changes have been motivated by the critical assessment of techniques for protein structure prediction (CASP) ligand-binding prediction experiment, which provided us with a framework to test the performance of firestar. URL: http://firedb.bioinfo.cnio.es/Php/FireStar.php.

FunFOLD: an improved automated method for the prediction of ligand binding residues using 3D models of proteins

Background

The accurate prediction of ligand binding residues from amino acid sequences is important for the automated functional annotation of novel proteins. In the previous two CASP experiments, the most successful methods in the function prediction category were those which used structural superpositions of 3D models and related templates with bound ligands in order to identify putative contacting residues. However, whilst most of this prediction process can be automated, visual inspection and manual adjustments of parameters, such as the distance thresholds used for each target, have often been required to prevent over prediction. Here we describe a novel method FunFOLD, which uses an automatic approach for cluster identification and residue selection. The software provided can easily be integrated into existing fold recognition servers, requiring only a 3D model and list of templates as inputs. A simple web interface is also provided allowing access to non-expert users. The method has been benchmarked against the top servers and manual prediction groups tested at both CASP8 and CASP9.

Results

The FunFOLD method shows a significant improvement over the best available servers and is shown to be competitive with the top manual prediction groups that were tested at CASP8. The FunFOLD method is also competitive with both the top server and manual methods tested at CASP9. When tested using common subsets of targets, the predictions from FunFOLD are shown to achieve a significantly higher mean Matthews Correlation Coefficient (MCC) scores and Binding-site Distance Test (BDT) scores than all server methods that were tested at CASP8. Testing on the CASP9 set showed no statistically significant separation in performance between FunFOLD and the other top server groups tested.

Conclusions

The FunFOLD software is freely available as both a standalone package and a prediction server, providing competitive ligand binding site residue predictions for expert and non-expert users alike. The software provides a new fully automated approach for structure based function prediction using 3D models of proteins.

Structure-based identification of catalytic residues

The identification of catalytic residues is an essential step in functional characterization of enzymes. We present a purely structural approach to this problem, which is motivated by the difficulty of evolution-based methods to annotate structural genomics targets that have few or no homologs in the databases. Our approach combines a state-of-the-art support vector machine (SVM) classifier with novel structural features that augment structural clues by spatial averaging and Z scoring. Special attention is paid to the class imbalance problem that stems from the overwhelming number of non-catalytic residues in enzymes compared to catalytic residues. This problem is tackled by: (1) optimizing the classifier to maximize a performance criterion that considers both Type I and Type II errors in the classification of catalytic and non-catalytic residues; (2) under-sampling non-catalytic residues before SVM training; and (3) during SVM training, penalizing errors in learning catalytic residues more than errors in learning non-catalytic residues. Tested on four enzyme datasets, one specifically designed by us to mimic the structural genomics scenario and three previously evaluated datasets, our structure-based classifier is never inferior to similar structure-based classifiers and comparable to classifiers that use both structural and evolutionary features. In addition to the evaluation of the performance of catalytic residue identification, we also present detailed case studies on three proteins. This analysis suggests that many false positive predictions may correspond to binding sites and other functional residues. A web server that implements the method, our own-designed database, and the source code of the programs are publicly available at http://www.cs.bgu.ac.il/∼meshi/functionPrediction.

The IntFOLD server: an integrated web resource for protein fold recognition, 3D model quality assessment, intrinsic disorder prediction, domain prediction and ligand binding site prediction

The IntFOLD server is a novel independent server that integrates several cutting edge methods for the prediction of structure and function from sequence. Our guiding principles behind the server development were as follows: (i) to provide a simple unified resource that makes our prediction software accessible to all and (ii) to produce integrated output for predictions that can be easily interpreted. The output for predictions is presented as a simple table that summarizes all results graphically via plots and annotated 3D models. The raw machine readable data files for each set of predictions are also provided for developers, which comply with the Critical Assessment of Methods for Protein Structure Prediction (CASP) data standards. The server comprises an integrated suite of five novel methods: nFOLD4, for tertiary structure prediction; ModFOLD 3.0, for model quality assessment; DISOclust 2.0, for disorder prediction; DomFOLD 2.0 for domain prediction; and FunFOLD 1.0, for ligand binding site prediction. Predictions from the IntFOLD server were found to be competitive in several categories in the recent CASP9 experiment. The IntFOLD server is available at the following web site: http://www.reading.ac.uk/bioinf/IntFOLD/.

Pharmacophore screening of the protein data bank for specific binding site chemistry

A simple computational approach was developed to screen the Protein Data Bank (PDB) for putative pockets possessing a specific binding site chemistry and geometry. The method employs two commonly used 3D screening technologies, namely identification of cavities in protein structures and pharmacophore screening of chemical libraries. For each protein structure, a pocket finding algorithm is used to extract potential binding sites containing the correct types of residues, which are then stored in a large SDF-formatted virtual library; pharmacophore filters describing the desired binding site chemistry and geometry are then applied to screen this virtual library and identify pockets matching the specified structural chemistry. As an example, this approach was used to screen all human protein structures in the PDB and identify sites having chemistry similar to that of known methyl-lysine binding domains that recognize chromatin methylation marks. The selected genes include known readers of the histone code as well as novel binding pockets that may be involved in epigenetic signaling. Putative allosteric sites were identified on the structures of TP53BP1, L3MBTL3, CHEK1, KDM4A, and CREBBP.

Predicting conserved protein motifs with Sub-HMMs

Background

Profile HMMs (hidden Markov models) provide effective methods for modeling the conserved regions of protein families. A limitation of the resulting domain models is the difficulty to pinpoint their much shorter functional sub-features, such as catalytically relevant sequence motifs in enzymes or ligand binding signatures of receptor proteins.

Results

To identify these conserved motifs efficiently, we propose a method for extracting the most information-rich regions in protein families from their profile HMMs. The method was used here to predict a comprehensive set of sub-HMMs from the Pfam domain database. Cross-validations with the PROSITE and CSA databases confirmed the efficiency of the method in predicting most of the known functionally relevant motifs and residues. At the same time, 46,768 novel conserved regions could be predicted. The data set also allowed us to link at least 461 Pfam domains of known and unknown function by their common sub-HMMs. Finally, the sub-HMM method showed very promising results as an alternative search method for identifying proteins that share only short sequence similarities.

Conclusions

Sub-HMMs extend the application spectrum of profile HMMs to motif discovery. Their most interesting utility is the identification of the functionally relevant residues in proteins of known and unknown function. Additionally, sub-HMMs can be used for highly localized sequence similarity searches that focus on shorter conserved features rather than entire domains or global similarities. The motif data generated by this study is a valuable knowledge resource for characterizing protein functions in the future.

From protein sequences to 3D-structures and beyond: the example of the UniProt knowledgebase

With the dramatic increase in the volume of experimental results in every domain of life sciences, assembling pertinent data and combining information from different fields has become a challenge. Information is dispersed over numerous specialized databases and is presented in many different formats. Rapid access to experiment-based information about well-characterized proteins helps predict the function of uncharacterized proteins identified by large-scale sequencing. In this context, universal knowledgebases play essential roles in providing access to data from complementary types of experiments and serving as hubs with cross-references to many specialized databases. This review outlines how the value of experimental data is optimized by combining high-quality protein sequences with complementary experimental results, including information derived from protein 3D-structures, using as an example the UniProt knowledgebase (UniProtKB) and the tools and links provided on its website ( http://www.uniprot.org/ ). It also evokes precautions that are necessary for successful predictions and extrapolations.

FragKB: structural and literature annotation resource of conserved peptide fragments and residues

Background

FragKB (Fragment Knowledgebase) is a repository of clusters of structurally similar fragments from proteins. Fragments are annotated with information at the level of sequence, structure and function, integrating biological descriptions derived from multiple existing resources and text mining.

Methodology

FragKB contains approximately 400,000 conserved fragments from 4,800 representative proteins from PDB. Literature annotations are extracted from more than 1,700 articles and are available for over 12,000 fragments. The underlying systematic annotation workflow of FragKB ensures efficient update and maintenance of this database. The information in FragKB can be accessed through a web interface that facilitates sequence and structural visualization of fragments together with known literature information on the consequences of specific residue mutations and functional annotations of proteins and fragment clusters. FragKB is accessible online at http://ubio.bioinfo.cnio.es/biotools/fragkb/.

Significance

The information presented in FragKB can be used for modeling protein structures, for designing novel proteins and for functional characterization of related fragments. The current release is focused on functional characterization of proteins through inspection of conservation of the fragments.

Cancer-associated mutations are preferentially distributed in protein kinase functional sites

Protein kinases are a superfamily involved in many crucial cellular processes, including signal transmission and regulation of cell cycle. As a consequence of this role, kinases have been reported to be associated with many types of cancer and are considered as potential therapeutic targets. We analyzed the distribution of pathogenic somatic point mutations (drivers) in the protein kinase superfamily with respect to their location in the protein, such as in structural, evolutionary, and functionally relevant regions. We find these driver mutations are more clearly associated with key protein features than other somatic mutations (passengers) that have not been directly linked to tumor progression. This observation fits well with the expected implication of the alterations in protein kinase function in cancer pathogenicity. To explain the relevance of the detected association of cancer driver mutations at the molecular level in the human kinome, we compare these with genetically inherited mutations (SNPs). We find that the subset of nonsynonymous SNPs that are associated to disease, but sufficiently mild to the point of being widespread in the population, tend to avoid those key protein regions, where they could be more detrimental for protein function. This tendency contrasts with the one detected for cancer associated-driver-mutations, which seems to be more directly implicated in the alteration of protein function. The detailed analysis of protein kinase groups and a number of relevant examples, confirm the relation between cancer associated-driver-mutations and key regions for protein kinase structure and function.

Protein interactions and ligand binding: from protein subfamilies to functional specificity

The divergence accumulated during the evolution of protein families translates into their internal organization as subfamilies, and it is directly reflected in the characteristic patterns of differentially conserved residues. These specifically conserved positions in protein subfamilies are known as "specificity determining positions" (SDPs). Previous studies have limited their analysis to the study of the relationship between these positions and ligand-binding specificity, demonstrating significant yet limited predictive capacity. We have systematically extended this observation to include the role of differential protein interactions in the segregation of protein subfamilies and explored in detail the structural distribution of SDPs at protein interfaces. Our results show the extensive influence of protein interactions in the evolution of protein families and the widespread association of SDPs with protein interfaces. The combined analysis of SDPs in interfaces and ligand-binding sites provides a more complete picture of the organization of protein families, constituting the necessary framework for a large scale analysis of the evolution of protein function.

Assessment of ligand binding residue predictions in CASP8

Here we detail the assessment process for the binding site prediction category of the eighth Critical Assessment of Protein Structure Prediction experiment (CASP8). Predictions were only evaluated for those targets that bound biologically relevant ligands and were assessed using the Matthews Correlation Coefficient. The results of the analysis clearly demonstrate that three predictors from two groups (Lee and Sternberg) stand out from the rest. A further two groups perform well over subsets of metal binding or nonmetal ligand binding targets. The best methods were able to make consistently reliable predictions based on model structures, though it was noticeable that the two targets that were not well predicted were also the hardest targets. The number of predictors that submitted new methods in this category was highly encouraging and suggests that current technology is at the level that experimental biochemists and structural biologists could benefit from what is clearly a growing field.

PepX: a structural database of non-redundant protein-peptide complexes

Although protein–peptide interactions are estimated to constitute up to 40% of all protein interactions, relatively little information is available for the structural details of these interactions. Peptide-mediated interactions are a prime target for drug design because they are predominantly present in signaling and regulatory networks. A reliable data set of nonredundant protein–peptide complexes is indispensable as a basis for modeling and design, but current data sets for protein–peptide interactions are often biased towards specific types of interactions or are limited to interactions with small ligands. In PepX (http://pepx.switchlab.org), we have designed an unbiased and exhaustive data set of all protein–peptide complexes available in the Protein Data Bank with peptide lengths up to 35 residues. In addition, these complexes have been clustered based on their binding interfaces rather than sequence homology, providing a set of structurally diverse protein–peptide interactions. The final data set contains 505 unique protein–peptide interface clusters from 1431 complexes. Thorough annotation of each complex with both biological and structural information facilitates searching for and browsing through individual complexes and clusters. Moreover, we provide an additional source of data for peptide design by annotating peptides with naturally occurring backbone variations using fragment clusters from the BriX database.

Genome and proteome annotation: organization, interpretation and integration

Recent years have seen a huge increase in the generation of genomic and proteomic data. This has been due to improvements in current biological methodologies, the development of new experimental techniques and the use of computers as support tools. All these raw data are useless if they cannot be properly analysed, annotated, stored and displayed. Consequently, a vast number of resources have been created to present the data to the wider community. Annotation tools and databases provide the means to disseminate these data and to comprehend their biological importance. This review examines the various aspects of annotation: type, methodology and availability. Moreover, it puts a special interest on novel annotation fields, such as that of phenotypes, and highlights the recent efforts focused on the integrating annotations.

LigASite--a database of biologically relevant binding sites in proteins with known apo-structures

Better characterization of binding sites in proteins and the ability to accurately predict their location and energetic properties are major challenges which, if addressed, would have many valuable practical applications. Unfortunately, reliable benchmark datasets of binding sites in proteins are still sorely lacking. Here, we present LigASite ('LIGand Attachment SITE'), a gold-standard dataset of binding sites in 550 proteins of known structures. LigASite consists exclusively of biologically relevant binding sites in proteins for which at least one apo- and one holo-structure are available. In defining the binding sites for each protein, information from all holo-structures is combined, considering in each case the quaternary structure defined by the PQS server. LigASite is built using simple criteria and is automatically updated as new structures become available in the PDB, thereby guaranteeing optimal data coverage over time. Both a redundant and a culled non-redundant version of the dataset is available at http://www.scmbb.ulb.ac.be/Users/benoit/LigASite. The website interface allows users to search the dataset by PDB identifiers, ligand identifiers, protein names or sequence, and to look for structural matches as defined by the CATH homologous superfamilies. The datasets can be downloaded from the website as Schema-validated XML files or comma-separated flat files.

firestar--prediction of functionally important residues using structural templates and alignment reliability

Here we present firestar, an expert system for predicting ligand-binding residues in protein structures. The server provides a method for extrapolating from the large inventory of functionally important residues organized in the FireDB database and adds information about the local conservation of potential-binding residues. The interface allows users to make queries by protein sequence or structure. The user can access pairwise and multiple alignments with structures that have relevant functionally important binding sites. The results are presented in a series of easy to read displays that allow users to compare binding residue conservation across homologous proteins. The binding site residues can also be viewed with molecular visualization tools. One feature of firestar is that it can be used to evaluate the biological relevance of small molecule ligands present in PDB structures. With the server it is easy to discern whether small molecule binding is conserved in homologous structures. We found this facility particularly useful during the recent assessment of CASP7 function prediction. Availability: http://firedb.bioinfo.cnio.es/Php/FireStar.php.

Assessment of predictions submitted for the CASP7 function prediction category

Here we present a full overview of the Critical Assessment of Protein Structure Prediction (CASP7) function prediction category. Predictions were submitted for Gene Ontology molecular function terms, Enzyme Commission numbers, and ligand binding site residues. The first two categories were difficult to assess because very little new functional information becomes available after the experiment. The majority of the known Gene Ontology terms and all the Enzyme Commission numbers were available a priori to predictors before the experiment, so prediction for these two categories was not blind. Nevertheless, for Gene Ontology terms we were able to demonstrate that some groups made better predictions than others. In the binding residue category, the predictors did not know in advance which ligands were bound and therefore blind evaluation was possible, but there were disappointingly few predictions in this category. After CASP 6 and 7 the need to organize a more effective blind function prediction category is obvious, even if it means focusing on binding site prediction as the only category that can be truly assessed in the CASP spirit.