Structure-based Classification Research Articles

Dynamics-Based Peptide–MHC Binding Optimization by a Convolutional Variational Autoencoder: A Use-Case Model for CASTELO

An unsolved challenge in the development of antigen-specific immunotherapies is determining the optimal antigens to target. Comprehension of antigen-major histocompatibility complex (MHC) binding is paramount toward achieving this goal. Here, we apply CASTELO, a combined machine learning-molecular dynamics (ML-MD) approach, to identify per-residue antigen binding contributions and then design novel antigens of increased MHC-II binding affinity for a type 1 diabetes-implicated system. We build upon a small-molecule lead optimization algorithm by training a convolutional variational autoencoder (CVAE) on MD trajectories of 48 different systems across four antigens and four HLA serotypes. We develop several new machine learning metrics including a structure-based anchor residue classification model as well as cluster comparison scores. ML-MD predictions agree well with experimental binding results and free energy perturbation-predicted binding affinities. Moreover, ML-MD metrics are independent of traditional MD stability metrics such as contact area and root-mean-square fluctuations (RMSF), which do not reflect binding affinity data. Our work supports the role of structure-based deep learning techniques in antigen-specific immunotherapy design.

Targeting more precisely: Improving sensitivity to EGFR inhibitors in NSCLC

The therapeutic landscape of epidermal growth factor receptor (EGFR)-mutation-positive non-small cell lung cancer (NSCLC) is continually evolving. A recent manuscript in Nature by Robichaux and colleagues1 reports on a structure-based classification of EGFR mutations to help predict sensitivities to EGFR inhibitors in NSCLC that may ultimately improve patient outcomes.

Structure-based classification of EGFR mutations informs inhibitor selection for lung cancer therapy.

EGFR oncogenic mutations predict sensitivity to EGFR inhibitors in NSCLC, but less is known about EGFR "variants of unknown significance." Using preclinical models, 3D structure analyses, and patient response data, Robichaux etal. show in Nature that mutations in structural regions of EGFR predict responses to different EGFR inhibitors.

RR3DD: an RNA global structure-based RNA three-dimensional structural classification database

ABSTRACT The three-dimensional (3D) structure of RNA usually plays an important role in the recognition with RNA-binding protein. Along with the discovering of RNAs, several RNA databases are developed to study the functions of RNA based on sequence, secondary structure, local 3D structural motif and global structure. Based on RNA function and structure, different RNAs are classified and stored in SCOR and DARTS, respectively. The classification of RNA structures is useful in RNA structure prediction and function annotation. However, the SCOR and DARTS are not updated any more. In this study, we present an RNA classification database RR3DD based on RNA fold with the global 3D structural similarity. The RR3DD includes 13,601 RNA chains from PDB and mmCIF format structures which are classified into 780 RNA folds. The RNA chains from PDB and mmCIF format structures are aligned and clustered into 675 and 220 RNA folds, respectively. By analysing the RNA structure in RR3DD, we find that there are 11 clusters with more than 50 members. These clusters include rRNAs, riboswitches, tRNAs and so on. By mapping RR3DD into Rfam, we found that some RNAs without annotation by Rfam can be annotated through structural alignment. For example, we analysed tRNAs and found that tRNA were successfully grouped in RR3DD for which Rfam did not classify them into one family. Finally, we provide a web interface of RR3DD offering functions of browsing RR3DD, annotating RNA 3D structure and finding templates for RNA homology modelling.

Structure-based classification of tauopathies.

The ordered assembly of tau protein into filaments characterizes several neurodegenerative diseases, which are called tauopathies. It was previously reported that, by cryo-electron microscopy, the structures of tau filaments from Alzheimer's disease1,2,Pick's disease3, chronic traumatic encephalopathy4 and corticobasal degeneration5 are distinct. Here we show that the structures of tau filaments from progressive supranuclear palsy (PSP) define a new three-layered fold. Moreover, the structures of tau filaments from globular glial tauopathy are similar to those from PSP. The tau filament fold of argyrophilic grain disease (AGD) differs, instead resembling the four-layered fold of corticobasal degeneration. The AGD fold is also observed in ageing-related tau astrogliopathy. Tau protofilament structures from inherited cases of mutations at positions +3 or +16 in intron 10 of MAPT (themicrotubule-associated protein taugene) are also identical to those from AGD, suggesting that relative overproduction of four-repeat tau can give rise to the AGD fold. Finally, the structures of tau filaments from cases of familial British dementia and familial Danish dementia are the same as those from cases of Alzheimer's disease and primary age-related tauopathy. These findings suggest a hierarchical classification of tauopathies on the basis of their filament folds, which complements clinical diagnosis and neuropathology and also allows the identification of new entities-as we show for a case diagnosed as PSP, but with filament structures that are intermediate between those of globular glial tauopathy and PSP.

Structure-based design and classifications of small molecules regulating the circadian rhythm period

Circadian rhythm is an important mechanism that controls behavior and biochemical events based on 24 h rhythmicity. Ample evidence indicates disturbance of this mechanism is associated with different diseases such as cancer, mood disorders, and familial delayed phase sleep disorder. Therefore, drug discovery studies have been initiated using high throughput screening. Recently the crystal structures of core clock proteins (CLOCK/BMAL1, Cryptochromes (CRY), Periods), responsible for generating circadian rhythm, have been solved. Availability of structures makes amenable core clock proteins to design molecules regulating their activity by using in silico approaches. In addition to that, the implementation of classification features of molecules based on their toxicity and activity will improve the accuracy of the drug discovery process. Here, we identified 171 molecules that target functional domains of a core clock protein, CRY1, using structure-based drug design methods. We experimentally determined that 115 molecules were nontoxic, and 21 molecules significantly lengthened the period of circadian rhythm in U2OS cells. We then performed a machine learning study to classify these molecules for identifying features that make them toxic and lengthen the circadian period. Decision tree classifiers (DTC) identified 13 molecular descriptors, which predict the toxicity of molecules with a mean accuracy of 79.53% using tenfold cross-validation. Gradient boosting classifiers (XGBC) identified 10 molecular descriptors that predict and increase in the circadian period length with a mean accuracy of 86.56% with tenfold cross-validation. Our results suggested that these features can be used in QSAR studies to design novel nontoxic molecules that exhibit period lengthening activity.

Structure-based classification predicts drug response in EGFR-mutant NSCLC

Epidermal growth factor receptor (EGFR) mutations typically occur in exons 18–21 and are established driver mutations in non-small cell lung cancer (NSCLC)1–3. Targeted therapies are approved for patients with ‘classical’ mutations and a small number of other mutations4–6. However, effective therapies have not been identified for additional EGFR mutations. Furthermore, the frequency and effects of atypical EGFR mutations on drug sensitivity are unknown1,3,7–10. Here we characterize the mutational landscape in 16,715 patients with EGFR-mutant NSCLC, and establish the structure–function relationship of EGFR mutations on drug sensitivity. We found that EGFR mutations can be separated into four distinct subgroups on the basis of sensitivity and structural changes that retrospectively predict patient outcomes following treatment with EGFR inhibitors better than traditional exon-based groups. Together, these data delineate a structure-based approach for defining functional groups of EGFR mutations that can effectively guide treatment and clinical trial choices for patients with EGFR-mutant NSCLC and suggest that a structure–function-based approach may improve the prediction of drug sensitivity to targeted therapies in oncogenes with diverse mutations.

Open reading frame 3 protein of hepatitis E virus: Multi-function protein with endless potential.

Hepatitis E virus (HEV), a fecal-orally transmitted foodborne viral pathogen, causes acute hepatitis in humans and is responsible for hepatitis E outbreaks worldwide. Since the identification of HEV as a zoonotic agent, this virus has been isolated from a variety of hosts with an ever-expanding host range. HEV-open reading frame (ORF) 3, the smallest ORF in HEV genomes, initially had been perceived as an unremarkable HEV accessory protein. However, as novel HEV-ORF3 function has been discovered that is related to the existence of a putative third virion structural form, referred to as “quasi-enveloped” HEV particles, HEV is challenging the conventional virion structure-based classification scheme, which assigns all viruses to two groups, “enveloped” or “non-enveloped”. In this review, we systematically describe recent progress that has identified multiple pathogenic roles of HEV-ORF3, including roles in HEV virion release, biogenesis of quasi-enveloped virus, regulation of the host innate immune response, and interference with host signaling pathways. In addition, implications of HEV-ORF3-associated quasi-enveloped virions are discussed to guide future development of improved vaccines against zoonotic HEV infection.

Learning chemistry: exploring the suitability of machine learning for the task of structure-based chemical ontology classification

Chemical data is increasingly openly available in databases such as PubChem, which contains approximately 110 million compound entries as of February 2021. With the availability of data at such scale, the burden has shifted to organisation, analysis and interpretation. Chemical ontologies provide structured classifications of chemical entities that can be used for navigation and filtering of the large chemical space. ChEBI is a prominent example of a chemical ontology, widely used in life science contexts. However, ChEBI is manually maintained and as such cannot easily scale to the full scope of public chemical data. There is a need for tools that are able to automatically classify chemical data into chemical ontologies, which can be framed as a hierarchical multi-class classification problem. In this paper we evaluate machine learning approaches for this task, comparing different learning frameworks including logistic regression, decision trees and long short-term memory artificial neural networks, and different encoding approaches for the chemical structures, including cheminformatics fingerprints and character-based encoding from chemical line notation representations. We find that classical learning approaches such as logistic regression perform well with sets of relatively specific, disjoint chemical classes, while the neural network is able to handle larger sets of overlapping classes but needs more examples per class to learn from, and is not able to make a class prediction for every molecule. Future work will explore hybrid and ensemble approaches, as well as alternative network architectures including neuro-symbolic approaches.

DbStRiPs: Database of structural repeats in proteins.

Recent interest in repeat proteins has arisen due to stable structural folds, high evolutionary conservation and repertoire of functions provided by these proteins. However, repeat proteins are poorly characterized because of high sequence variation between repeating units and structure-based identification and classification of repeats is desirable. Using a robust network-based pipeline, manual curation and Kajava's structure-based classification schema, we have developed a database of tandem structural repeats, Database of Structural Repeats in Proteins (DbStRiPs). A unique feature of this database is that available knowledge on sequence repeat families is incorporated by mapping Pfam classification scheme onto structural classification. Integration of sequence and structure-based classifications help in identifying different functional groups within the same structural subclass, leading to refinement in the annotation of repeat proteins. Analysis of complete Protein Data Bank revealed 16,472 repeat annotations in 15,141 protein chains, one previously uncharacterized novel protein repeat family (PRF), named left-handed beta helix, and 33 protein repeat clusters (PRCs). Based on their unique structural motif, ~79% of these repeat proteins are classified in one of the 14 PRFs or 33 PRCs, and the remaining are grouped as unclassified repeat proteins. Each repeat protein is provided with a detailed annotation in DbStRiPs that includes start and end boundaries of repeating units, copy number, secondary and tertiary structure view, repeat class/subclass, disease association, MSA of repeating units and cross-references to various protein pattern databases, human protein atlas and interaction resources. DbStRiPs provides easy search and download options to high-quality annotations of structural repeat proteins (URL: http://bioinf.iiit.ac.in/dbstrips/).

Subcapularis tendon tear: A structure-based arthroscopic description

IntroductionSubscapularis (SSC) tendons differ from supraspinatus tendons, although both have similar histologic structure comprising two layers with distinct collagen fiber organization. HypothesisThe partial/full-thickness tear classification for the supraspinatus based on tendon structure can be applied to the subscapularis tendon on objective criteria. Material and methodsThe present study used 100 films of arthroscopic rotator cuff repair involving SSC lesion. Lesions were reported on 3 objective criteria: horizontal superior tendon edge visibility, lesser tuberosity bone exposure, and lateral tendon edge visibility. Combining the three distinguishes deep, superficial or interstitial partial tear versus full-thickness tear. Degree of retraction was also noted. ResultsForty-six of the 73 partial lesions involved the deep articular layer, which was often retracted, but conserving the horizontal superior tendon edge and thus misleadingly suggesting SSC integrity; 23 were interstitial, without detachment from the lesser tuberosity; 4 involved only the superficial layer. Full-thickness tears were always retracted, with loss of horizontal superior tendon edge, visibility of the lateral tendon edge and presence of comma sign. Inter- and intra-observer reproducibility was satisfactory. Discussion – conclusionLike in superior cuff tear, a structure-based classification can be made of SSC lesions on objective criteria. Level of evidenceIV.

CoV3D: a database of high resolution coronavirus protein structures.

SARS-CoV-2, the etiologic agent of COVID-19, exemplifies the general threat to global health posed by coronaviruses. The urgent need for effective vaccines and therapies is leading to a rapid rise in the number of high resolution structures of SARS-CoV-2 proteins that collectively reveal a map of virus vulnerabilities. To assist structure-based design of vaccines and therapeutics against SARS-CoV-2 and other coronaviruses, we have developed CoV3D, a database and resource for coronavirus protein structures, which is updated on a weekly basis. CoV3D provides users with comprehensive sets of structures of coronavirus proteins and their complexes with antibodies, receptors, and small molecules. Integrated molecular viewers allow users to visualize structures of the spike glycoprotein, which is the major target of neutralizing antibodies and vaccine design efforts, as well as sets of spike-antibody complexes, spike sequence variability, and known polymorphisms. In order to aid structure-based design and analysis of the spike glycoprotein, CoV3D permits visualization and download of spike structures with modeled N-glycosylation at known glycan sites, and contains structure-based classification of spike conformations, generated by unsupervised clustering. CoV3D can serve the research community as a centralized reference and resource for spike and other coronavirus protein structures, and is available at: https://cov3d.ibbr.umd.edu.

Evolutionary and functional classification of the CARF domain superfamily, key sensors in prokaryotic antivirus defense.

CRISPR-associated Rossmann Fold (CARF) and SMODS-associated and fused to various effector domains (SAVED) are key components of cyclic oligonucleotide-based antiphage signaling systems (CBASS) that sense cyclic oligonucleotides and transmit the signal to an effector inducing cell dormancy or death. Most of the CARFs are components of a CBASS built into type III CRISPR-Cas systems, where the CARF domain binds cyclic oligoA (cOA) synthesized by Cas10 polymerase-cyclase and allosterically activates the effector, typically a promiscuous ribonuclease. Additionally, this signaling pathway includes a ring nuclease, often also a CARF domain (either the sensor itself or a specialized enzyme) that cleaves cOA and mitigates dormancy or death induction. We present a comprehensive census of CARF and SAVED domains in bacteria and archaea, and their sequence- and structure-based classification. There are 10 major families of CARF domains and multiple smaller groups that differ in structural features, association with distinct effectors, and presence or absence of the ring nuclease activity. By comparative genome analysis, we predict specific functions of CARF and SAVED domains and partition the CARF domains into those with both sensor and ring nuclease functions, and sensor-only ones. Several families of ring nucleases functionally associated with sensor-only CARF domains are also predicted.

Struc2gauss: Structural role preserving network embedding via Gaussian embedding

Network embedding (NE) is playing a principal role in network mining, due to its ability to map nodes into efficient low-dimensional embedding vectors. However, two major limitations exist in state-of-the-art NE methods: role preservation and uncertainty modeling. Almost all previous methods represent a node into a point in space and focus on local structural information, i.e., neighborhood information. However, neighborhood information does not capture global structural information and point vector representation fails in modeling the uncertainty of node representations. In this paper, we propose a new NE framework, struc2gauss, which learns node representations in the space of Gaussian distributions and performs network embedding based on global structural information. struc2gauss first employs a given node similarity metric to measure the global structural information, then generates structural context for nodes and finally learns node representations via Gaussian embedding. Different structural similarity measures of networks and energy functions of Gaussian embedding are investigated. Experiments conducted on real-world networks demonstrate that struc2gauss effectively captures global structural information while state-of-the-art network embedding methods fail to, outperforms other methods on the structure-based clustering and classification task and provides more information on uncertainties of node representations.

Quaternary structural diversity in eukaryotic DNA polymerases: monomeric to multimeric form.

Sixteen eukaryotic DNA polymerases have been identified and studied so far. Based on the sequence similarity of the catalytic subunits of DNA polymerases, these have been classified into four A, B, X and Y families except PrimPol, which belongs to the AEP family. The quaternary structure of these polymerases also varies depending upon whether they are composed of one or more subunits. Therefore, in this review, we used a quaternary structure-based classification approach to group DNA polymerases as either monomeric or multimeric and highlighted functional significance of their accessory subunits. Additionally, we have briefly summarized various DNA polymerase discoveries from a historical perspective, emphasized unique catalytic mechanism of each DNA polymerase and highlighted recent advances in understanding their cellular functions.

Unraveling virus relationships by structure-based phylogenetic classification.

Delineation of the intricacies of protein function from macromolecular structure constitutes a continual obstacle in the study of cell and pathogen biology. Structure-based phylogenetic analysis has emerged as a powerful tool for addressing this challenge, allowing the detection and quantification of conserved architectural properties between proteins, including those with low or no detectable sequence homology. With a focus on viral protein structure, we highlight how a number of investigations have utilized this powerful method to infer common functionality and ancestry.

IStable 2.0: Predicting protein thermal stability changes by integrating various characteristic modules

Protein mutations can lead to structural changes that affect protein function and result in disease occurrence. In protein engineering, drug design or and optimization industries, mutations are often used to improve protein stability or to change protein properties while maintaining stability. To provide possible candidates for novel protein design, several computational tools for predicting protein stability changes have been developed. Although many prediction tools are available, each tool employs different algorithms and features. This can produce conflicting prediction results that make it difficult for users to decide upon the correct protein design. Therefore, this study proposes an integrated prediction tool, iStable 2.0, which integrates 11 sequence-based and structure-based prediction tools by machine learning and adds protein sequence information as features. Three coding modules are designed for the system, an Online Server Module, a Stand-alone Module and a Sequence Coding Module, to improve the prediction performance of the previous version of the system. The final integrated structure-based classification model has a higher Matthews correlation coefficient than that of the single prediction tool (0.708 vs 0.547, respectively), and the Pearson correlation coefficient of the regression model likewise improves from 0.669 to 0.714. The sequence-based model not only successfully integrates off-the-shelf predictors but also improves the Matthews correlation coefficient of the best single prediction tool by at least 0.161, which is better than the individual structure-based prediction tools. In addition, both the Sequence Coding Module and the Stand-alone Module maintain performance with only a 5% decrease of the Matthews correlation coefficient when the integrated online tools are unavailable. iStable 2.0 is available at http://ncblab.nchu.edu.tw/iStable2.

Typological features and prospective lines of development of roadside service facilities in Belarus

Introduction: the methods of roadside service network development implemented in Belarus were identified. Structure, architectural and spatial solutions, and general functional zoning of a service facility territory were considered for the roadside service facilities located along Belarus main motor roads. The study aims at the revelation of typological features of roadside service network elements in Belarus and their structure-based classification as well as the development of the main provisions on the formation of these facilities based on the current situation analysis and prospective lines of building up a service system in the areas of influence of national importance roads.
 
 
 Materials and methods: the work was carried out through systematization, structural and comparative analyses, theoretical synthesis of the data obtained from the detailed analysis of literature and statistical sources, field surveys of roadside service facilities located along the national-importance roads in Belarus. Likewise design estimation documentation for the constructed facilities was applied. During the field surveys, photo fixation and graph-analytic analysis methods were used.
 
 
 Results: the article suggests the classification of roadside service facilities, reveals the methods of network development, considers functional zoning and offers a concept of prospective formation of the facilities in Belarus.
 
 
 Conclusions: theoretical and methodological foundations of the facility formation were developed, which allows accomplishing more reasonable and objective decisions regarding the development of the roadside service network. The developed classification allows defining new approaches to the formation of the network individual elements. The suggested concept of a modular system for the prospective formation of the roadside service in Belarus can form a basis for the initial build-up and temporary transformation of the facilities and the system as a whole and can be used to develop regulatory requirements for their planning structure and architectural and spatial solution.

Structural and Phylogenetic Diversity of Anaerobic Carbon-Monoxide Dehydrogenases.

Anaerobic Ni-containing carbon-monoxide dehydrogenases (Ni-CODHs) catalyze the reversible conversion between carbon monoxide and carbon dioxide as multi-enzyme complexes responsible for carbon fixation and energy conservation in anaerobic microbes. However, few biochemically characterized model enzymes exist, with most Ni-CODHs remaining functionally unknown. Here, we performed phylogenetic and structure-based Ni-CODH classification using an expanded dataset comprised of 1942 non-redundant Ni-CODHs from 1375 Ni-CODH-encoding genomes across 36 phyla. Ni-CODHs were divided into seven clades, including a novel clade. Further classification into 24 structural groups based on sequence analysis combined with structural prediction revealed diverse structural motifs for metal cluster formation and catalysis, including novel structural motifs potentially capable of forming metal clusters or binding metal ions, indicating Ni-CODH diversity and plasticity. Phylogenetic analysis illustrated that the metal clusters responsible for intermolecular electron transfer were drastically altered during evolution. Additionally, we identified novel putative Ni-CODH-associated proteins from genomic contexts other than the Wood–Ljungdahl pathway and energy converting hydrogenase system proteins. Network analysis among the structural groups of Ni-CODHs, their associated proteins and taxonomies revealed previously unrecognized gene clusters for Ni-CODHs, including uncharacterized structural groups with putative metal transporters, oxidoreductases, or transcription factors. These results suggested diversification of Ni-CODH structures adapting to their associated proteins across microbial genomes.

Structure-Based Classification and Anti-Cancer Effects of Plant Metabolites

A variety of malignant cancers affect the global human population. Although a wide variety of approaches to cancer treatment have been studied and used clinically (surgery, radiotherapy, chemotherapy, and immunotherapy), the toxic side effects of cancer therapies have a negative impact on patients and impede progress in conquering cancer. Plant metabolites are emerging as new leads for anti-cancer drug development. This review summarizes these plant metabolites with regard to their structures and the types of cancer against which they show activity, organized by the organ or tissues in which each cancer forms. This information will be helpful for understanding the current state of knowledge of the anti-cancer effects of various plant metabolites against major types of cancer for the further development of novel anti-cancer drugs.