Structure-based Classification Research Articles

Flavoprotein oxidases: classification and applications

This review provides an overview of oxidases that utilise a flavin cofactor for catalysis. This class of oxidative flavoenzymes has shown to harbour a large number of biotechnologically interesting enzymes. Applications range from their use as biocatalysts for the synthesis of pharmaceutical compounds to the integration in biosensors. Through the recent developments in genome sequencing, the number of newly discovered oxidases is steadily growing. Recent progress in the field of flavoprotein oxidase discovery and the obtained biochemical knowledge on these enzymes are reviewed. Except for a structure-based classification of known flavoprotein oxidases, also their potential in recent biotechnological applications is discussed.

A Pluralist Approach to Extension: The Role of Materiality in Scientific Practice for the Reference of Natural Kind Terms

This article argues for a different outlook on the concept of extension, especially for the reference of general terms in scientific practice. Scientific realist interpretations of the two predominant theories of meaning, namely Descriptivism and Causal Theory, contend that a stable cluster of descriptions or an initial baptism fixes the extension of a general term such as a natural kind term. This view in which the meaning of general terms is presented as monosemantic and the referents as stable, homogeneous, and unchangeable, however, does not reflect the various practices involved in the investigation of research materials and the related application of general terms in scientific practice. By drawing on the taxonomic diversity, particularly of structure-based classifications in chemical databases, this article illustrates the limited utility of such a concept of extension. Research materials often exhibit a plurality of material dimensions that, within different research contexts, allow for various and often equally significant taxonomic demarcations. In light of this, the extension of a general term cannot be uniquely determined by a supposedly independent nature of the referent but is relative to the model context under which the materials are investigated. This significance and plurality of the model context, I claim, needs to be mirrored in an account of meaning that is supposed to reflect scientific reality. On this account, the aim of this article is to present an alternative perspective on the concept of extension to accommodate the diverse material practices that determine the application of general terms in science.

Structural characterization of Pandoraea pnomenusa B-356 biphenyl dioxygenase reveals features of potent polychlorinated biphenyl-degrading enzymes.

The oxidative degradation of biphenyl and polychlorinated biphenyls (PCBs) is initiated in Pandoraea pnomenusa B-356 by biphenyl dioxygenase (BPDOB356). BPDOB356, a heterohexameric (αβ)3 Rieske oxygenase (RO), catalyzes the insertion of dioxygen with stereo- and regioselectivity at the 2,3-carbons of biphenyl, and can transform a broad spectrum of PCB congeners. Here we present the X-ray crystal structures of BPDOB356 with and without its substrate biphenyl 1.6-Å resolution for both structures. In both cases, the Fe(II) has five ligands in a square pyramidal configuration: H233 Nε2, H239 Nε2, D386 Oδ1 and Oδ2, and a single water molecule. Analysis of the active sites of BPDOB356 and related ROs revealed structural features that likely contribute to the superior PCB-degrading ability of certain BPDOs. First, the active site cavity readily accommodates biphenyl with minimal conformational rearrangement. Second, M231 was predicted to sterically interfere with binding of some PCBs, and substitution of this residue yielded variants that transform 2,2′-dichlorobiphenyl more effectively. Third, in addition to the volume and shape of the active site, residues at the active site entrance also apparently influence substrate preference. Finally, comparison of the conformation of the active site entrance loop among ROs provides a basis for a structure-based classification consistent with a phylogeny derived from amino acid sequence alignments.

Characterization of the Type III sulfide:quinone oxidoreductase from Caldivirga maquilingensis and its membrane binding

Sulfide:quinone oxidoreductases (SQRs) are ubiquitous enzymes which have multiple roles: sulfide detoxification, energy generation by providing electrons to respiratory or photosynthetic electron transfer chains, and sulfide homeostasis. A recent structure-based classification defines 6 groups of putative SQRs (I–VI), and representatives of all but group III have been confirmed to have sulfide oxidase activity. In the current work, we report the first characterization of a predicted group III SQR from Caldivirga maquilingensis, and confirm that this protein is a sulfide oxidase. The gene encoding the enzyme was cloned, and the protein was expressed in E. coli and purified. The enzyme oxidizes sulfide using decylubiquinone as an electron acceptor, and is inhibited by aurachin C and iodoacetamide. Analysis of the amino acid sequence indicates that the C. maquilingensis SQR has two amphiphilic helices at the C-terminus but lacks any transmembrane helices. This suggests that C. maquilingensis SQR interacts with the membrane surface and that the interactions are mediated by the C-terminal amphiphilic helices. Mutations within the last C-terminal amphiphilic helix resulted in a water-soluble form of the enzyme which, remarkably, retains full SQR activity using decylubiquinone as the electron acceptor. Mutations at one position, L379, also located in the C-terminal amphiphilic helix, inactivated the enzyme by preventing the interaction with decylubiquinone. It is concluded that the C-terminal amphiphilic helix is important for membrane binding and for forming part of the pathway providing access of the quinone substrate to the protein-bound flavin at the enzyme active site.

Predicting Ligand Binding Residues and Functional Sites Using Multipositional Correlations with Graph Theoretic Clustering and Kernel CCA

We present a new computational method for predicting ligand binding residues and functional sites in protein sequences. These residues and sites tend to be not only conserved, but also exhibit strong correlation due to the selection pressure during evolution in order to maintain the required structure and/or function. To explore the effect of correlations among multiple positions in the sequences, the method uses graph theoretic clustering and kernel-based canonical correlation analysis (kCCA) to identify binding and functional sites in protein sequences as the residues that exhibit strong correlation between the residues’ evolutionary characterization at the sites and the structure-based functional classification of the proteins in the context of a functional family. The results of testing the method on two well-curated data sets show that the prediction accuracy as measured by Receiver Operating Characteristic (ROC) scores improves significantly when multipositional correlations are accounted for.

Structure-based classification and ontology in chemistry.

BackgroundRecent years have seen an explosion in the availability of data in the chemistry domain. With this information explosion, however, retrieving relevant results from the available information, and organising those results, become even harder problems. Computational processing is essential to filter and organise the available resources so as to better facilitate the work of scientists. Ontologies encode expert domain knowledge in a hierarchically organised machine-processable format. One such ontology for the chemical domain is ChEBI. ChEBI provides a classification of chemicals based on their structural features and a role or activity-based classification. An example of a structure-based class is 'pentacyclic compound' (compounds containing five-ring structures), while an example of a role-based class is 'analgesic', since many different chemicals can act as analgesics without sharing structural features. Structure-based classification in chemistry exploits elegant regularities and symmetries in the underlying chemical domain. As yet, there has been neither a systematic analysis of the types of structural classification in use in chemistry nor a comparison to the capabilities of available technologies.ResultsWe analyze the different categories of structural classes in chemistry, presenting a list of patterns for features found in class definitions. We compare these patterns of class definition to tools which allow for automation of hierarchy construction within cheminformatics and within logic-based ontology technology, going into detail in the latter case with respect to the expressive capabilities of the Web Ontology Language and recent extensions for modelling structured objects. Finally we discuss the relationships and interactions between cheminformatics approaches and logic-based approaches.ConclusionSystems that perform intelligent reasoning tasks on chemistry data require a diverse set of underlying computational utilities including algorithmic, statistical and logic-based tools. For the task of automatic structure-based classification of chemical entities, essential to managing the vast swathes of chemical data being brought online, systems which are capable of hybrid reasoning combining several different approaches are crucial. We provide a thorough review of the available tools and methodologies, and identify areas of open research.

Cluster analysis of the DrugBank chemical space using molecular quantum numbers

DrugBank (>6000 approved and experimental drugs) was analyzed using molecular quantum numbers (MQNs), which are 42 integer value descriptors of molecular structure counting atoms, bonds, polar groups and topological features. Principal component analysis of MQN-space showed that drugs differ mostly by size (PC1, 67% variance) and structural rigidity and polarity (PC2, 18% variance). Twenty-eight groups of target specific drugs were recovered by proximity sorting in MQN-space as efficiently as by substructure fingerprint (SF) similarity, but in different order allowing for lead-hopping relationships not seen in SF similarity. Clustering by MQN- or SF-similarity produced very different types of clusters. Each of the 28 drug groups spread over different clusters in both MQN- and SF-clustering, and most clusters contained drugs from different target specific groups, showing that structure-based classifications only partially overlap with bioactivity. An MQN-browsable version of DrugBank is available at www.gdb.unibe.ch.

Camps 2.0: Exploring the sequence and structure space of prokaryotic, eukaryotic, and viral membrane proteins

Structural bioinformatics of membrane proteins is still in its infancy, and the picture of their fold space is only beginning to emerge. Because only a handful of three-dimensional structures are available, sequence comparison and structure prediction remain the main tools for investigating sequence-structure relationships in membrane protein families. Here we present a comprehensive analysis of the structural families corresponding to α-helical membrane proteins with at least three transmembrane helices. The new version of our CAMPS database (CAMPS 2.0) covers nearly 1300 eukaryotic, prokaryotic, and viral genomes. Using an advanced classification procedure, which is based on high-order hidden Markov models and considers both sequence similarity as well as the number of transmembrane helices and loop lengths, we identified 1353 structurally homogeneous clusters roughly corresponding to membrane protein folds. Only 53 clusters are associated with experimentally determined three-dimensional structures, and for these clusters CAMPS is in reasonable agreement with structure-based classification approaches such as SCOP and CATH. We therefore estimate that ∼1300 structures would need to be determined to provide a sufficient structural coverage of polytopic membrane proteins. CAMPS 2.0 is available at http://webclu.bio.wzw.tum.de/CAMPS2.0/.

Prototype semantic infrastructure for automated small molecule classification and annotation in lipidomics

BackgroundThe development of high-throughput experimentation has led to astronomical growth in biologically relevant lipids and lipid derivatives identified, screened, and deposited in numerous online databases. Unfortunately, efforts to annotate, classify, and analyze these chemical entities have largely remained in the hands of human curators using manual or semi-automated protocols, leaving many novel entities unclassified. Since chemical function is often closely linked to structure, accurate structure-based classification and annotation of chemical entities is imperative to understanding their functionality.ResultsAs part of an exploratory study, we have investigated the utility of semantic web technologies in automated chemical classification and annotation of lipids. Our prototype framework consists of two components: an ontology and a set of federated web services that operate upon it. The formal lipid ontology we use here extends a part of the LiPrO ontology and draws on the lipid hierarchy in the LIPID MAPS database, as well as literature-derived knowledge. The federated semantic web services that operate upon this ontology are deployed within the Semantic Annotation, Discovery, and Integration (SADI) framework. Structure-based lipid classification is enacted by two core services. Firstly, a structural annotation service detects and enumerates relevant functional groups for a specified chemical structure. A second service reasons over lipid ontology class descriptions using the attributes obtained from the annotation service and identifies the appropriate lipid classification. We extend the utility of these core services by combining them with additional SADI services that retrieve associations between lipids and proteins and identify publications related to specified lipid types. We analyze the performance of SADI-enabled eicosanoid classification relative to the LIPID MAPS classification and reflect on the contribution of our integrative methodology in the context of high-throughput lipidomics.ConclusionsOur prototype framework is capable of accurate automated classification of lipids and facile integration of lipid class information with additional data obtained with SADI web services. The potential of programming-free integration of external web services through the SADI framework offers an opportunity for development of powerful novel applications in lipidomics. We conclude that semantic web technologies can provide an accurate and versatile means of classification and annotation of lipids.

An evaluation of structure-based toxicity classification and TTC approach for assessing safety of existing industrial chemicals with JECDB

An evaluation of structure-based toxicity classification and TTC approach for assessing safety of existing industrial chemicals with JECDB

Chemical ontologies: what are they, what are they for and what are the challenges

Ontologies encode human knowledge in computationally accessible forms. They are designed to narrow the gap between the knowledge of human experts and the functionality available in computer systems, by expressing expert knowledge in a manner computers can manipulate and reason over. With the ever-growing deluge of data in modern scientific domains, researchers need intelligent tools able to filter out irrelevant and automatically organise relevant information into meaningful categories. The Chemoinformatics and Metabolism team at the EBI is developing chemical ontologies for structure-based chemical classification, role or bioactivity-based chemical classification, and chemical information entities such as descriptors and algorithms. Our ontologies provide collections of names and synonyms which are useful for text mining, stable identifiers which are essential to semantic integration of data, and a semantically rich encoding of many aspects of the chemical domain. But for such ontologies to be maximally useful for diverse users and interoperable with other ontologies in the scientific domain, similar-sounding things have to be disentangled in our language and our ontology. Recent work addresses the distinguishing of structures from chemicals [1], and of bioactivity from drug uses [2]. Ontologies are backed by logical formalisms such as the Web Ontology Language, OWL. One of the challenges of chemical ontology is representing complex chemical structures in the underlying formalism. Cyclic structures prove particularly challenging for logicbased representation. Recent research in our group investigated the inclusion of chemical graphs in OWL [3]. Integrating chemoinformatics tools with chemical ontology is the subject of ongoing research.

Structure-Function Analysis of Grass Clip Serine Protease Involved in Drosophila Toll Pathway Activation

Grass is a clip domain serine protease (SP) involved in a proteolytic cascade triggering the Toll pathway activation of Drosophila during an immune response. Epistasic studies position it downstream of the apical protease ModSP and upstream of the terminal protease Spaetzle-processing enzyme. Here, we report the crystal structure of Grass zymogen. We found that Grass displays a rather deep active site cleft comparable with that of proteases of coagulation and complement cascades. A key distinctive feature is the presence of an additional loop (75-loop) in the proximity of the activation site localized on a protruding loop. All biochemical attempts to hydrolyze the activation site of Grass failed, strongly suggesting restricted access to this region. The 75-loop is thus proposed to constitute an original mechanism to prevent spontaneous activation. A comparison of Grass with clip serine proteases of known function involved in analogous proteolytic cascades allowed us to define two groups, according to the presence of the 75-loop and the conformation of the clip domain. One group (devoid of the 75-loop) contains penultimate proteases whereas the other contains terminal proteases. Using this classification, Grass appears to be a terminal protease. This result is evaluated according to the genetic data documenting Grass function.

A New Structure-based Classification of Gram-positive Bacteriocins

Bacteriocins are ribosomally-synthesized peptides or proteins produced by a wide range of bacteria. The antimicrobial activity of this group of natural substances against foodborne pathogenic and spoilage bacteria has raised considerable interest for their application in food preservation. Classifying these bacteriocins in well defined classes according to their biochemical properties is a major step towards characterizing these anti-infective peptides and understanding their mode of action. Actually, the chosen criteria for bacteriocins' classification lack consistency and coherence. So, various classification schemes of bacteriocins resulted various levels of contradiction and sorting inefficiencies leading to bacteriocins belonging to more than one class at the same time and to a general lack of classification of many bacteriocins. Establishing a coherent and adequate classification scheme for these bacteriocins is sought after by several researchers in the field. It is not straightforward to formulate an efficient classification scheme that encompasses all of the existing bacteriocins. In the light of the structural data, here we revisit the previously proposed contradictory classification and we define new structure-based sequence fingerprints that support a subdivision of the bacteriocins into 12 groups. The paper lays down a resourceful and consistent classification approach that resulted in classifying more than 70% of bacteriocins known to date and with potential to identify distinct classes for the remaining unclassified bacteriocins. Identified groups are characterized by the presence of highly conserved short amino acid motifs. Furthermore, unclassified bacteriocins are expected to form an identified group when there will be sufficient sequences.

T-RMSD: A Fine-grained, Structure-based Classification Method and its Application to the Functional Characterization of TNF Receptors

This study addresses the relation between structural and functional similarity in proteins. We introduce a novel method named tree based on root mean square deviation (T-RMSD), which uses distance RMSD (dRMSD) variations to build fine-grained structure-based classifications of proteins. The main improvement of the T-RMSD over similar methods, such as Dali, is its capacity to produce the equivalent of a bootstrap value for each cluster node. We validated our approach on two domain families studied extensively for their role in many biological and pathological pathways: the small GTPase RAS superfamily and the cysteine-rich domains (CRDs) associated with the tumor necrosis factor receptors (TNFRs) family. Our analysis showed that T-RMSD is able to automatically recover and refine existing classifications. In the case of the small GTPase ARF subfamily, T-RMSD can distinguish GTP- from GDP-bound states, while in the case of CRDs it can identify two new subgroups associated with well defined functional features (ligand binding and formation of ligand pre-assembly complex). We show how hidden Markov models (HMMs) can be built on these new groups and propose a methodology to use these models simultaneously in order to do fine-grained functional genomic annotation without known 3D structures. T-RMSD, an open source freeware incorporated in the T-Coffee package, is available online.

Fingerprint-based detection of acute aquatic toxicity

In this work we show the effectiveness of 2D structural fingerprints in the prediction of aquatic toxicity of chemical compounds, creating a self-contained system for structure-based aquatic toxicity classification. Using the data from the U.S. Environmental Protection Agency Fat Head Minnow (EPA-FHM) dataset [1] we build a nonlinear RBF SVM [2] classifier that distinguishes acutely toxic compounds from less toxic compounds, loosely according to the criterion stipulated by the E.U. Reach legislation [3]. The classifier achieves up to 86% accuracy in leave-one-out validation using 580 of the dataset’s 614 compounds. This performance is comparable with models built from the same dataset using more sophisticated molecular descriptors, such as AutoMEP and Sterimol descriptors [4]. We apply our classification model to predict the aquatic toxicity of 3M compounds in the MMsINC database [5]. Furthermore, we create a linear SVM model using the same technique and apply it to the MMsINC data, with the additional integration of the EXPLAIN system [6] which allows us to show which structural features are responsible for the model classifying a molecule as less toxic or acutely toxic.

Functional States of Homooligomers: Insights from the Evolution of Glycosyltransferases

Glycosylation is an important aspect of epigenetic regulation. Glycosyltransferase is a key enzyme in the biosynthesis of glycans, which glycosylates more than half of all proteins in eukaryotes and is involved in a wide range of biological processes. It has been suggested previously that homooligomerization in glycosyltransferases and other proteins might be crucial for their function. In this study, we explore functional homooligomeric states of glycosyltransferases in various organisms, trace their evolution, and perform comparative analyses to find structural features that can mediate or disrupt the formation of different homooligomers. First, we make a structure-based classification of the diverse superfamily of glycosyltransferases and confirm that the majority of the structures are indeed clustered into the GT-A or GT-B folds. We find that homooligomeric glycosyltransferases appear to be as ancient as monomeric glycosyltransferases and go back in evolution to the last universal common ancestor (LUCA). Moreover, we show that interface residues have significant bias to be gapped out or unaligned in the monomers, implying that they might represent features crucial for oligomer formation. Structural analysis of these features reveals that the majority of them represent loops, terminal regions, and helices, indicating that these secondary-structure elements mediate the formation of glycosyltransferases' homooligomers and directly contribute to the specific binding. We also observe relatively short protein regions that disrupt the homodimer interactions, although such cases are rare. These results suggest that relatively small structural changes in the nonconserved regions may contribute to the formation of different functional oligomeric states and might be important in regulation of enzyme activity through homooligomerization.

Update of the LIPID MAPS comprehensive classification system for lipids

In 2005, the International Lipid Classification and Nomenclature Committee under the sponsorship of the LIPID MAPS Consortium developed and established a "Comprehensive Classification System for Lipids" based on well-defined chemical and biochemical principles and using an ontology that is extensible, flexible, and scalable. This classification system, which is compatible with contemporary databasing and informatics needs, has now been accepted internationally and widely adopted. In response to considerable attention and requests from lipid researchers from around the globe and in a variety of fields, the comprehensive classification system has undergone significant revisions over the last few years to more fully represent lipid structures from a wider variety of sources and to provide additional levels of detail as necessary. The details of this classification system are reviewed and updated and are presented here, along with revisions to its suggested nomenclature and structure-drawing recommendations for lipids.

Machine learning approach for structure-based zeolite classification

Application of knowledge discovery methods to crystal structure databases is an emerging research area of materials science that is playing an important role in facilitating data analysis. This study is aimed at combining computational geometry methods with machine learning algorithms for classification of inorganic solid materials of known structure. Zeolite crystals are used for the pilot study where a model based on the topology is developed for classification of the compound by mineral name and by zeolite framework type. The topological descriptors are derived from the Delaunay tessellation for 220 zeolites contained in the inorganic crystal structure database. This zeolite-structure-predictor (ZSP) is trained for classifying this set of selected zeolite crystals into 22 different types of minerals and into 13 framework types. The ZSP is based on the random forest algorithm and contains attributes of Delaunay simplex properties such as tetrahedrality index, frequency of simplex occurrence, and site occupation probability. The ZSP is able to obtain classification in this multitude of classes with more than 81% of correctly classified instances based on framework type. The model shows that the classification into framework types is superior, and that the classification into mineral names is not structurally unique.

Molecular Architecture of the Major Histocompatibility Complex Class I-binding Site of Ly49 Natural Killer Cell Receptors

Natural killer (NK) cells play a vital role in the detection and destruction of virally infected and tumor cells during innate immune responses. The highly polymorphic Ly49 family of NK receptors regulates NK cell function by sensing major histocompatibility complex class I (MHC-I) molecules on target cells. Despite the determination of two Ly49-MHC-I complex structures, the molecular features of Ly49 receptors that confer specificity for particular MHC-I alleles have not been identified. To understand the functional architecture of Ly49-binding sites, we determined the crystal structures of Ly49C and Ly49G and completed refinement of the Ly49C-H-2K(b) complex. This information, combined with mutational analysis of Ly49A, permitted a structure-based classification of Ly49s that we used to dissect the binding site into three distinct regions, each having different roles in MHC recognition. One region, located at the center of the binding site, has a similar structure across the Ly49 family and mediates conserved interactions with MHC-I that contribute most to binding. However, the preference of individual Ly49s for particular MHC-I molecules is governed by two regions that flank the central region and are structurally more variable. One of the flanking regions divides Ly49s into those that recognize both H-2D and H-2K versus only H-2D ligands, whereas the other discriminates among H-2D or H-2K alleles. The modular design of Ly49-binding sites provides a framework for predicting the MHC-binding specificity of Ly49s that have not been characterized experimentally.

A Structural Model of the Membrane‐Bound Aromatic Prenyltransferase UbiA from E. coli

The membrane-bound enzyme 4-hydroxybenzoic acid oligoprenyltransferase (ubiA) from E. coli is crucial for the production of ubiquinone, the essential electron carrier in prokaryotic and eukaryotic organisms. On the basis of previous modeling analyses, amino acids identified as important in two putative active sites (1 and 2) were selectively mutated. All mutants but one lost their ability to form geranylated hydroxybenzoate, irrespective of their being from active site 1 or 2. This suggests either that the two active sites are interrelated or that they are in fact only one site. With the aid of the experimental results and a new structure-based classification of prenylating enzymes, a relevant 3D model could be developed by threading. The new model explains the substrate specificities and is in complete agreement with the results of site-directed mutagenesis. The high similarity of the active fold of UbiA-transferase to that of 5-epi-aristolochene synthase (Nicotiana tabacum), despite a low homology, allows a hypothesis on a convergent evolution of these enzymes to be formed.