Search

Respiratory tract infections (RTIs) have a significant impact on global health, especially among children and the elderly. The key bacterial pathogens Streptococcus pneumoniae, Haemophilus influenzae, Klebsiella pneumoniae, Staphylococcus aureus and non-fermenting Gram Negative bacteria such as Acinetobacter baumannii and Pseudomonas aeruginosa are most commonly associated with RTIs. These bacterial pathogens have evolved a diverse array of resistance mechanisms through horizontal gene transfer, often mediated by mobile genetic elements and environmental acquisition. Treatment failures are primarily due to antimicrobial resistance and inadequate bacterial engagement, which necessitates the development of alternative treatment strategies. To overcome this, our review mainly focuses on different virulence mechanisms and their resulting pathogenicity, highlighting different therapeutic interventions to combat resistance. To prevent the antimicrobial resistance crisis, we also focused on leveraging the application of artificial intelligence and machine learning to manage RTIs. Integrative approaches combining mechanistic insights are crucial for addressing the global challenge of antimicrobial resistance in respiratory infections. [ABSTRACT FROM AUTHOR]

Beta turns, in which the protein backbone abruptly changes direction over four amino acid residues, are the most common type of protein secondary structure after alpha helices and beta sheets and play key structural and functional roles. Previous work has produced classification systems for turn geometry at multiple levels of precision, but these operate in backbone dihedral‐angle (Ramachandran) space, and the absence of a local Euclidean‐space coordinate system and structural alignment for turns, or of any systematic Euclidean‐space characterization of turn backbone shape, presents challenges for the visualization, comparison and analysis of the wide range of turn conformations and the design of turns and the structures that incorporate them. This work derives a turn‐local coordinate system that implicitly aligns turns, together with a set of geometric descriptors that characterize the bulk BB shapes of turns and describe modes of structural variation not explicitly captured by existing systems. These modes are shown to be meaningful by the demonstration of clear relationships between descriptor values and the electrostatic energy of the beta‐turn H‐bond, the overrepresentations of key side‐chain motifs, and the structural contexts of turns. Geometric turn descriptors complement Ramachandran‐space classifications, and they can be used to select turn structures for compatibility with particular side‐chain interactions or contexts. Potential applications include protein design and other tasks in which an enhanced Euclidean‐space characterization of turns may improve understanding or performance. The web‐based tools ExploreTurns, MapTurns, and ProfileTurn, available at www.betaturn.com, incorporate turn‐local coordinates and turn descriptors and demonstrate their utility. [ABSTRACT FROM AUTHOR]

KRAS belongs to a family of small GTPases that act as binary switches upstream of several signalling cascades, controlling proliferation and survival of cells. Mutations in KRAS drive oncogenesis, especially in pancreatic, lung, and colorectal cancers (CRC). Although historic attempts at targeting mutant KRAS with small molecule inhibitors have proven challenging, there are recent successes with the G12C, and G12D mutations. However, clinically important RAS mutations such as G12V, G13D, Q61L, and A146T, remain elusive drug targets, and insights to their structural landscape is of critical importance to develop novel, and effective therapeutic concepts. We present a fully open, P-loop exposing conformer of KRAS G13D by X-ray crystallography at 1.4–2.4 Å resolution in Mg2+-free phosphate and malonate buffers. The G13D conformer has the switch-I region displaced in an upright position leaving the catalytic core fully exposed. To prove that this state is druggable, we developed a P-loop-targeting monoclonal antibody (mAb). The mAb displayed high-affinity binding to G13D and was shown using high resolution fluorescence microscopy to be spontaneously taken up by G13D-mutated HCT 116 cells (human CRC derived) by macropinocytosis. The mAb inhibited KRAS signalling in phosphoproteomic and genomic studies. Taken together, the data propose novel druggable space of G13D that is reachable in the cellular context. It is our hope that these findings will stimulate attempts to drug this fully open state G13D conformer using mAbs or other modalities. [ABSTRACT FROM AUTHOR]

The results of tertiary structure assessment at CASP15 are reported. For the first time, recognizing the outstanding performance of AlphaFold 2 (AF2) at CASP14, all single-chain predictions were assessed together, irrespective of whether a template was available. At CASP15, there was no single stand-out group, with most of the best-scoring groups-led by PEZYFoldings, UM-TBM, and Yang Server-employing AF2 in one way or another. Many top groups paid special attention to generating deep Multiple Sequence Alignments (MSAs) and testing variant MSAs, thereby allowing them to successfully address some of the hardest targets. Such difficult targets, as well as lacking templates, were typically proteins with few homologues. Local divergence between prediction and target correlated with localization at crystal lattice or chain interfaces, and with regions exhibiting high B-factor factors in crystal structure targets, and should not necessarily be considered as representing error in the prediction. However, analysis of exposed and buried side chain accuracy showed room for improvement even in the latter. Nevertheless, a majority of groups produced high-quality predictions for most targets, which are valuable for experimental structure determination, functional analysis, and many other tasks across biology. These include those applying methods similar to those used to generate major resources such as the AlphaFold Protein Structure Database and the ESM Metagenomic atlas: the confidence estimates of the former were also notably accurate.

Many research groups and institutions have created a variety of databases curating experimental and predicted data related to protein-ligand binding. The landscape of available databases is dynamic, with new databases emerging and established databases becoming defunct. Here, we review the current state of databases that contain binding pockets and protein-ligand binding interactions. We have compiled a list of such databases, fifty-three of which are currently available for use. We discuss variation in how binding pockets are defined and summarize pocket-finding methods. We organize the fifty-three databases into subgroups based on goals and contents, and describe standard use cases. We also illustrate that pockets within the same protein are characterized differently across different databases. Finally, we assess critical issues of sustainability, accessibility and redundancy.

Abstract The type VI secretion system (T6SS) is an important mediator of microbe–microbe and microbe–host interactions. Gram-negative bacteria use the T6SS to inject T6SS effectors (T6Es), which are usually proteins with toxic activity, into neighboring cells. Antibacterial effectors have cognate immunity proteins that neutralize self-intoxication. Here, we applied novel structural bioinformatic tools to perform systematic discovery and functional annotation of T6Es and their cognate immunity proteins from a dataset of 17,920 T6SS-encoding bacterial genomes. Using structural clustering, we identified 517 putative T6E families, outperforming sequence-based clustering. We developed a logistic regression model to reliably quantify protein–protein interaction of new T6E-immunity pairs, yielding candidate immunity proteins for 231 out of the 517 T6E families. We used sensitive structure-based annotation which yielded functional annotations for 51% of the T6E families, again outperforming sequence-based annotation. Next, we validated four novel T6E-immunity pairs using basic experiments in E. coli. In particular, we showed that the Pfam domain DUF3289 is a homolog of Colicin M and that DUF943 acts as its cognate immunity protein. Furthermore, we discovered a novel T6E that is a structural homolog of SleB, a lytic transglycosylase, and identified a specific glutamate that acts as its putative catalytic residue. Overall, this study applies novel structural bioinformatic tools to T6E-immunity pair discovery, and provides an extensive database of annotated T6E-immunity pairs.

The type VI secretion system (T6SS) is an important mediator of microbe–microbe and microbe–host interactions. Gram-negative bacteria use the T6SS to inject T6SS effectors (T6Es), which are usually proteins with toxic activity, into neighboring cells. Antibacterial effectors have cognate immunity proteins that neutralize self-intoxication. Here, we applied novel structural bioinformatic tools to perform systematic discovery and functional annotation of T6Es and their cognate immunity proteins from a dataset of 17,920 T6SS-encoding bacterial genomes. Using structural clustering, we identified 517 putative T6E families, outperforming sequence-based clustering. We developed a logistic regression model to reliably quantify protein–protein interaction of new T6E-immunity pairs, yielding candidate immunity proteins for 231 out of the 517 T6E families. We used sensitive structure-based annotation which yielded functional annotations for 51% of the T6E families, again outperforming sequence-based annotation. Next, we validated four novel T6E-immunity pairs using basic experiments in E. coli. In particular, we showed that the Pfam domain DUF3289 is a homolog of Colicin M and that DUF943 acts as its cognate immunity protein. Furthermore, we discovered a novel T6E that is a structural homolog of SleB, a lytic transglycosylase, and identified a specific glutamate that acts as its putative catalytic residue. Overall, this study applies novel structural bioinformatic tools to T6E-immunity pair discovery, and provides an extensive database of annotated T6E-immunity pairs. Synopsis: Structural bioinformatic tools were utilized for the discovery of novel specialized Type VI Secretion System (T6SS) effectors and their cognate immunity proteins, highlighting their utility over standard sequence-based tools. The effector predictions were supported by experimental results. Structural clustering provided better compression of effectors than sequence-based methods, with 517 structural clusters representing the structure space of specialized effectors in Proteobacteria. The ipTM score from Alphafold-multimer was used as a reliable and quantitative measure for predicting candidate immunity proteins in 231 out of 517 effector clusters. Annotations were provided for 265 out of the 517 specialized effector domain families using fast and sensitive searches with Foldseek, expanding capabilities beyond Pfam-based annotation alone. Four putative effectors were demonstrated to be toxic to Escherichia coli, with co-expression of cognate immunity proteins neutralizing their toxicity. Structural bioinformatic tools were utilized for the discovery of novel specialized Type VI Secretion System (T6SS) effectors and their cognate immunity proteins, highlighting their utility over standard sequence-based tools. The effector predictions were supported by experimental results. [ABSTRACT FROM AUTHOR]

Motivation Tools for pairwise alignments between 3D structures of proteins are of fundamental importance for structural biology and bioinformatics, enabling visual exploration of evolutionary and functional relationships. However, the absence of a user-friendly, browser-based tool for creating alignments and visualizing them at both 1D sequence and 3D structural levels makes this process unnecessarily cumbersome. Results We introduce a novel pairwise structure alignment tool (rcsb.org/alignment) that seamlessly integrates into the RCSB Protein Data Bank (RCSB PDB) research-focused RCSB.org web portal. Our tool and its underlying application programming interface (alignment.rcsb.org) empowers users to align several protein chains with a reference structure by providing access to established alignment algorithms (FATCAT, CE, TM-align, or Smith–Waterman 3D). The user-friendly interface simplifies parameter setup and input selection. Within seconds, our tool enables visualization of results in both sequence (1D) and structural (3D) perspectives through the RCSB PDB RCSB.org Sequence Annotations viewer and Mol* 3D viewer, respectively. Users can effortlessly compare structures deposited in the PDB archive alongside more than a million incorporated Computed Structure Models coming from the ModelArchive and AlphaFold DB. Moreover, this tool can be used to align custom structure data by providing a link/URL or uploading atomic coordinate files directly. Importantly, alignment results can be bookmarked and shared with collaborators. By bridging the gap between 1D sequence and 3D structures of proteins, our tool facilitates deeper understanding of complex evolutionary relationships among proteins through comprehensive sequence and structural analyses. Availability and implementation The alignment tool is part of the RCSB PDB research-focused RCSB.org web portal and available at rcsb.org/alignment. Programmatic access is available via alignment.rcsb.org. Frontend code has been published at github.com/rcsb/rcsb-pecos-app. Visualization is powered by the open-source Mol* viewer (github.com/molstar/molstar and github.com/molstar/rcsb-molstar) plus the Sequence Annotations in 3D Viewer (github.com/rcsb/rcsb-saguaro-3d). [ABSTRACT FROM AUTHOR]

Reliable and accurate methods of estimating the accuracy of predicted protein models are vital to understanding their respective utility. Discerning how the quaternary structure conforms can significantly improve our collective understanding of cell biology, systems biology, disease formation, and disease treatment. Accurately determining the quality of multimeric protein models is still computationally challenging, as the space of possible conformations is significantly larger when proteins form in complex with one another. Here, we present EGG (energy and graph-based architectures) to assess the accuracy of predicted multimeric protein models. We implemented message-passing and transformer layers to infer the overall fold and interface accuracy scores of predicted multimeric protein models. When evaluated with CASP15 targets, our methods achieved promising results against single model predictors: fourth and third place for determining the highest-quality model when estimating overall fold accuracy and overall interface accuracy, respectively, and first place for determining the top three highest quality models when estimating both overall fold accuracy and overall interface accuracy. [ABSTRACT FROM AUTHOR]

Activating signal co-integrator complex 1 (ASCC1) acts with ASCC-ALKBH3 complex in alkylation damage responses. ASCC1 uniquely combines two evolutionarily ancient domains: nucleotide-binding K-Homology (KH) (associated with regulating splicing, transcriptional, and translation) and two-histidine phosphodiesterase (PDE; associated with hydrolysis of cyclic nucleotide phosphate bonds). Germline mutations link loss of ASCC1 function to spinal muscular atrophy with congenital bone fractures 2 (SMABF2). Herein analysis of The Cancer Genome Atlas (TCGA) suggests ASCC1 RNA overexpression in certain tumors correlates with poor survival, Signatures 29 and 3 mutations, and genetic instability markers. We determined crystal structures of Alvinella pompejana (Ap) ASCC1 and Human (Hs) PDE domain revealing high-resolution details and features conserved over 500 million years of evolution. Extending our understanding of the KH domain Gly-X-X-Gly sequence motif, we define a novel structural Helix-Clasp-Helix (HCH) nucleotide binding motif and show ASCC1 sequence-specific binding to CGCG-containing RNA. The V-shaped PDE nucleotide binding channel has two His-Φ-Ser/Thr-Φ (HXT) motifs (Φ being hydrophobic) positioned to initiate cyclic phosphate bond hydrolysis. A conserved atypical active-site histidine torsion angle implies a novel PDE substrate. Flexible active site loop and arginine-rich domain linker appear regulatory. Small-angle X-ray scattering (SAXS) revealed aligned KH-PDE RNA binding sites with limited flexibility in solution. Quantitative evolutionary bioinformatic analyses of disease and cancer-associated mutations support implied functional roles for RNA binding, phosphodiesterase activity, and regulation. Collective results inform ASCC1's roles in transactivation and alkylation damage responses, its targeting by structure-based inhibitors, and how ASCC1 mutations may impact inherited disease and cancer. [ABSTRACT FROM AUTHOR]

Protein tyrosine phosphatase 1B (PTP1B) is a validated therapeutic target for obesity, diabetes, and certain types of cancer. In particular, allosteric inhibitors hold potential for therapeutic use, but an incomplete understanding of conformational dynamics and allostery in this protein has hindered their development. Here, we interrogate solution dynamics and allosteric responses in PTP1B using high‐resolution hydrogen‐deuterium exchange mass spectrometry (HDX‐MS), an emerging and powerful biophysical technique. Using HDX‐MS, we obtain a detailed map of backbone amide exchange that serves as a proxy for the solution dynamics of apo PTP1B, revealing several flexible loops interspersed among more constrained and rigid regions within the protein structure, as well as local regions that exchange faster than expected from their secondary structure and solvent accessibility. We demonstrate that our HDX rate data obtained in solution adds value to estimates of conformational heterogeneity derived from a pseudo‐ensemble constructed from ~200 crystal structures of PTP1B. Furthermore, we report HDX‐MS maps for PTP1B with active‐site versus allosteric small‐molecule inhibitors. These maps suggest distinct and widespread effects on protein dynamics relative to the apo form, including changes in locations distal (>35 Å) from the respective ligand binding sites. These results illuminate that allosteric inhibitors of PTP1B can induce unexpected changes in dynamics that extend beyond the previously understood allosteric network. Together, our data suggest a model of BB3 allostery in PTP1B that combines conformational restriction of active‐site residues with compensatory liberation of distal residues that aid in entropic balancing. Overall, our work showcases the potential of HDX‐MS for elucidating aspects of protein conformational dynamics and allosteric effects of small‐molecule ligands and highlights the potential of integrating HDX‐MS alongside other complementary methods, such as room‐temperature X‐ray crystallography, NMR spectroscopy, and molecular dynamics simulations, to guide the development of new therapeutics. [ABSTRACT FROM AUTHOR]

Protein-nucleic acid (PNA) binding plays critical roles in the transcription, translation, regulation, and three-dimensional organization of the genome. Structural models of proteins bound to nucleic acids (NA) provide insights into the chemical, electrostatic, and geometric properties of the protein structure that give rise to NA binding but are scarce relative to models of unbound proteins. We developed a deep learning approach for predicting PNA binding given the unbound structure of a protein that we call PNAbind. Our method utilizes graph neural networks to encode the spatial distribution of physicochemical and geometric properties of protein structures that are predictive of NA binding. Using global physicochemical encodings, our models predict the overall binding function of a protein, and using local encodings, they predict the location of individual NA binding residues. Our models can discriminate between specificity for DNA or RNA binding, and we show that predictions made on computationally derived protein structures can be used to gain mechanistic understanding of chemical and structural features that determine NA recognition. Binding site predictions were validated against benchmark datasets, achieving AUROC scores in the range of 0.92–0.95. We applied our models to the HIV-1 restriction factor APOBEC3G and showed that our model predictions are consistent with and help explain experimental RNA binding data. [ABSTRACT FROM AUTHOR]

This work is an overview of deep machine learning methods aimed at predicting binding sites in protein structures. Several classes of methods are selected: prediction of binding sites for small molecules, proteins, and nucleic acids. For each class, various approaches to prediction are considered (prediction of binding atoms, residues, surfaces, pockets). Specifics of feature selection and neural network architectures inherent to each class and approach are highlighted, and an attempt is made to explain these specifics and foresee the further direction of their development. [ABSTRACT FROM AUTHOR]

Brucellosis is a zoonotic disease with significant economic and healthcare costs. Despite the eradication efforts, the disease persists. Vaccines prevent disease in animals while antibiotics cure humans with limitations. This study aims to design vaccines and drugs for brucellosis in animals and humans, using protein modeling, epitope prediction, and molecular docking of the target proteins (BvrR, OMP25, and OMP31). Tertiary structure models of three target proteins were constructed and assessed using RMSD, TM-score, C-score, Z-score, and ERRAT. The best models selected from AlphaFold and I-TASSER due to their superior performance according to CASP 12 – CASP 15 were chosen for further analysis. The motif analysis of best models using MotifFinder revealed two, five, and five protein binding motifs, however, the Motif Scan identified seven, six, and eight Post-Translational Modification sites (PTMs) in the BvrR, OMP25, and OMP31 proteins, respectively. Dominant B cell epitopes were predicted at (44–63, 85–93, 126–137, 193–205, and 208–237), (26–46, 52–71, 98–114, 142–155, and 183–200), and (29–45, 58–82, 119–142, 177–198, and 222–251) for the three target proteins. Additionally, cytotoxic T lymphocyte epitopes were detected at (173–181, 189–197, and 202–210), (61–69, 91–99, 159–167, and 181–189), and (3–11, 24–32, 167–175, and 216–224), while T helper lymphocyte epitopes were displayed at (39–53, 57–65, 150–158, 163–171), (79–87, 95–108, 115–123, 128–142, and 189–197), and (39–47, 109–123, 216–224, and 245–253), for the respective target protein. Furthermore, structure-based virtual screening of the ZINC and DrugBank databases using the docking MOE program was followed by ADMET analysis. The best five compounds of the ZINC database revealed docking scores ranged from (− 16.8744 to − 15.1922), (− 16.0424 to − 14.1645), and (− 14.7566 to − 13.3222) for the BvrR, OMP25, and OMP31, respectively. These compounds had good ADMET parameters and no cytotoxicity, while DrugBank compounds didn't meet Lipinski's rule criteria. Therefore, the five selected compounds from the ZINC20 databases may fulfill the pharmacokinetics and could be considered lead molecules for potentially inhibiting Brucella's proteins. [ABSTRACT FROM AUTHOR]

In this review article, we explore the transformative impact of deep learning (DL) on structural bioinformatics, emphasizing its pivotal role in a scientific revolution driven by extensive data, accessible toolkits and robust computing resources. As big data continue to advance, DL is poised to become an integral component in healthcare and biology, revolutionizing analytical processes. Our comprehensive review provides detailed insights into DL, featuring specific demonstrations of its notable applications in bioinformatics. We address challenges tailored for DL, spotlight recent successes in structural bioinformatics and present a clear exposition of DL—from basic shallow neural networks to advanced models such as convolution, recurrent, artificial and transformer neural networks. This paper discusses the emerging use of DL for understanding biomolecular structures, anticipating ongoing developments and applications in the realm of structural bioinformatics. [ABSTRACT FROM AUTHOR]

Diabetes mellitus (DM) represents a problem for the healthcare system worldwide. DM has very serious complications such as blindness, kidney failure, and cardiovascular disease. In addition to the very bad socioeconomic impacts, it influences patients and their families and communities. The global costs of DM and its complications are huge and expected to rise by the year 2030. DM is caused by genetic and environmental risk factors. Genetic testing will aid in early diagnosis and identification of susceptible individuals or populations using ATP-sensitive potassium (KATP) channels present in different tissues such as the pancreas, myocardium, myocytes, and nervous tissues. The channels respond to different concentrations of blood sugar, stimulation by hormones, or ischemic conditions. In pancreatic cells, they regulate the secretion of insulin and glucagon. Mutations in the KCNJ11 gene that encodes the Kir6.2 protein (a major constituent of KATP channels) were reported to be associated with Type 2 DM, neonatal diabetes mellitus (NDM), and maturity-onset diabetes of the young (MODY). Kir6.2 harbors binding sites for ATP and phosphatidylinositol 4,5-diphosphate (PIP2). The ATP inhibits the KATP channel, while the (PIP2) activates it. A Kir6.2 mutation at tyrosine330 (Y330) was demonstrated to reduce ATP inhibition and predisposes to NDM. In this study, we examined the effect of mutations on the Kir6.2 structure using bioinformatics tools and molecular dynamic simulations (SIFT, PolyPhen, SNAP2, PANTHER, PhD&SNP, SNP&Go, I-Mutant, MuPro, MutPred, ConSurf, HOPE, and GROMACS). Our results indicated that M199R, R201H, R206H, and Y330H mutations influence Kir6.2 structure and function and therefore may cause DM. We conclude that MD simulations are useful techniques to predict the effects of mutations on protein structure. In addition, the M199R, R201H, R206H, and Y330H variant in the Kir6.2 protein may be associated with DM. These results require further verification in protein–protein interactions, Kir6.2 function, and case-control studies. [ABSTRACT FROM AUTHOR]

Glutamate transporters play key roles in nervous physiology by modulating excitatory neurotransmitter levels, when malfunctioning, involving in a wide range of neurological and physiological disorders. However, integral transmembrane proteins including the glutamate transporters remain notoriously difficult to study, due to their localization within the cell membrane. Here we present the structural bioinformatics studies of glutamate transporters and their water-soluble variants generated through QTY-code, a protein design strategy based on systematic amino acid substitutions. These include 2 structures determined by X-ray crystallography, cryo-EM, and 6 predicted by AlphaFold2, and their predicted water-soluble QTY variants. In the native structures of glutamate transporters, transmembrane helices contain hydrophobic amino acids such as leucine (L), isoleucine (I), and phenylalanine (F). To design water-soluble variants, these hydrophobic amino acids are systematically replaced by hydrophilic amino acids, namely glutamine (Q), threonine (T) and tyrosine (Y). The QTY variants exhibited water-solubility, with four having identical isoelectric focusing points (pI) and the other four having very similar pI. We present the superposed structures of the native glutamate transporters and their water-soluble QTY variants. The superposed structures displayed remarkable similarity with RMSD 0.528Å-2.456Å, despite significant protein transmembrane sequence differences (41.1%—>53.8%). Additionally, we examined the differences of hydrophobicity patches between the native glutamate transporters and their QTY variants. Upon closer inspection, we discovered multiple natural variations of L->Q, I->T, F->Y and Q->L, T->I, Y->F in these transporters. Some of these natural variations were benign and the remaining were reported in specific neurological disorders. We further investigated the characteristics of hydrophobic to hydrophilic substitutions in glutamate transporters, utilizing variant analysis and evolutionary profiling. Our structural bioinformatics studies not only provided insight into the differences between the hydrophobic helices and hydrophilic helices in the glutamate transporters, but they are also expected to stimulate further study of other water-soluble transmembrane proteins. [ABSTRACT FROM AUTHOR]

Abstract Background Machine learning (ML) has a rich history in structural bioinformatics, and modern approaches, such as deep learning, are revolutionizing our knowledge of the subtle relationships between biomolecular sequence, structure, function, dynamics and evolution. As with any advance that rests upon statistical learning approaches, the recent progress in biomolecular sciences is enabled by the availability of vast volumes of sufficiently-variable data. To be useful, such data must be well-structured, machine-readable, intelligible and manipulable. These and related requirements pose challenges that become especially acute at the computational scales typical in ML. Furthermore, in structural bioinformatics such data generally relate to protein three-dimensional (3D) structures, which are inherently more complex than sequence-based data. A significant and recurring challenge concerns the creation of large, high-quality, openly-accessible datasets that can be used for specific training and benchmarking tasks in ML pipelines for predictive modeling projects, along with reproducible splits for training and testing. Results Here, we report ‘Prop3D’, a platform that allows for the creation, sharing and extensible reuse of libraries of protein domains, featurized with biophysical and evolutionary properties that can range from detailed, atomically-resolved physicochemical quantities (e.g., electrostatics) to coarser, residue-level features (e.g., phylogenetic conservation). As a community resource, we also supply a ‘Prop3D-20sf’ protein dataset, obtained by applying our approach to CATH . We have developed and deployed the Prop3D framework, both in the cloud and on local HPC resources, to systematically and reproducibly create comprehensive datasets via the Highly Scalable Data Service ( HSDS ). Our datasets are freely accessible via a public HSDS instance, or they can be used with accompanying Python wrappers for popular ML frameworks. Conclusion Prop3D and its associated Prop3D-20sf dataset can be of broad utility in at least three ways. Firstly, the Prop3D workflow code can be customized and deployed on various cloud-based compute platforms, with scalability achieved largely by saving the results to distributed HDF5 files via HSDS . Secondly, the linked Prop3D-20sf dataset provides a hand-crafted, already-featurized dataset of protein domains for 20 highly-populated CATH families; importantly, provision of this pre-computed resource can aid the more efficient development (and reproducible deployment) of ML pipelines. Thirdly, Prop3D-20sf’s construction explicitly takes into account (in creating datasets and data-splits) the enigma of ‘data leakage’, stemming from the evolutionary relationships between proteins.

Abstract Proteins are the key molecular machines that orchestrate all biological processes of the cell. Most proteins fold into three-dimensional shapes that are critical for their function. Studying the 3D shape of proteins can inform us of the mechanisms that underlie biological processes in living cells and can have practical applications in the study of disease mutations or the discovery of novel drug treatments. Here, we review the progress made in sequence-based prediction of protein structures with a focus on applications that go beyond the prediction of single monomer structures. This includes the application of deep learning methods for the prediction of structures of protein complexes, different conformations, the evolution of protein structures and the application of these methods to protein design. These developments create new opportunities for research that will have impact across many areas of biomedical research.

O-GlcNAcylation is a nutrient-sensing post-translational modification process. This cycling process involves two primary proteins: the O-linked N-acetylglucosamine transferase (OGT) catalysing the addition, and the glycoside hydrolase OGA (O-GlcNAcase) catalysing the removal of the O-GlCNAc moiety on nucleocytoplasmic proteins. This process is necessary for various critical cellular functions. The O-linked N-acetylglucosamine transferase (OGT) gene produces the OGT protein. Several studies have shown the overexpression of this protein to have biological implications in metabolic diseases like cancer and diabetes mellitus (DM). This study retrieved 159 SNPs with clinical significance from the SNPs database. We probed the functional effects, stability profile, and evolutionary conservation of these to determine their fit for this research. We then identified 7 SNPs (G103R, N196K, Y228H, R250C, G341V, L367F, and C845S) with predicted deleterious effects across the four tools used (PhD-SNPs, SNPs&Go, PROVEAN, and PolyPhen2). Proceeding with this, we used ROBETTA, a homology modelling tool, to model the proteins with these point mutations and carried out a structural bioinformatics method– molecular docking– using the Glide model of the Schrodinger Maestro suite. We used a previously reported inhibitor of OGT, OSMI-1, as the ligand for these mutated protein models. As a result, very good binding affinities and interactions were observed between this ligand and the active site residues within 4Å of OGT. We conclude that these mutation points may be used for further downstream analysis as drug targets for treating diabetes mellitus. [ABSTRACT FROM AUTHOR]

Centrosomes and cilia are microtubule-based superstructures vital for cell division, signaling, and motility. The once thought hollow lumen of their microtubule core structures was recently found to hold a rich meshwork of microtubule inner proteins (MIPs). To address the outstanding question of how distinct MIPs evolved to recognize microtubule inner surfaces, we applied computational sequence analyses, structure predictions, and experimental validation to uncover evolutionarily conserved microtubule- and MIP-binding modules named NWE, SNYG, and ELLEn, and PYG and GFG-repeat by their signature motifs. These modules intermix with MT-binding DM10-modules and Mn-repeats in 24 Chlamydomonas and 33 human proteins. The modules molecular characteristics provided keys to identify elusive cross-species homologs, hitherto unknown human MIP candidates, and functional properties for seven protein subfamilies, including the microtubule seam-binding NWE and ELLEn families. Our work defines structural innovations that underpin centriole and axoneme assembly and demonstrates that MIPs co-evolved with centrosomes and cilia. The inside surface of microtubules contains so-called microtubule inner proteins, but little is known about their identity. Here the authors use bioinformatics to identify structural motifs within this class of proteins and potential new members. [ABSTRACT FROM AUTHOR]

Proteins are the key molecular machines that orchestrate all biological processes of the cell. Most proteins fold into three-dimensional shapes that are critical for their function. Studying the 3D shape of proteins can inform us of the mechanisms that underlie biological processes in living cells and can have practical applications in the study of disease mutations or the discovery of novel drug treatments. Here, we review the progress made in sequence-based prediction of protein structures with a focus on applications that go beyond the prediction of single monomer structures. This includes the application of deep learning methods for the prediction of structures of protein complexes, different conformations, the evolution of protein structures and the application of these methods to protein design. These developments create new opportunities for research that will have impact across many areas of biomedical research. Predicting protein structure from sequence information has been a long-standing challenge. This Review discusses recent developments and applications of deep learning-based methods for protein structure prediction and design. [ABSTRACT FROM AUTHOR]

As an important macro element for all living cells, phosphorus is essential in agricultural production systems and is required in large quantities by elite varieties of crops to maintain yields. Approximately 70% of the worldwide cultivated land suffers from phosphorous deficiency, and it has recently been estimated that the worldwide phosphorous resources will be shattered by the end of this century, thereby increasing the need to develop phosphorus-efficient crops. A greater understanding of how plants can maintain yield with lower phosphorous availability is highly desirable to both breeders and farmers. For this research, we focused on the phosphorusstarvation tolerance 1 (PSTOL1) gene, which is known to be involved in enhancing early root growth, thereby enabling plants to acquire more phosphorus and other nutrients and enhancing grain yield in phosphorus-deficient soil. As there is no reported structure and function of the PSTOL1 gene, this project involves a distinct set of opportunities and challenges and requires different approaches to model the interaction between PSTOL1 and the gene phosphorous uptake 1 (PUP1). This article covers the modeling, docking, and simulation of PSTOL1 to check the protein stability and its behavior over time. The physiochemical properties were ascertained, a phylogenetic tree was constructed to find the evolutionary relationship, and the conserved domains were analyzed for functional annotation. This study reports that the advancement of the PSTOL1-mediated phosphorous uptake 1 (PUP1) signaling cascade using structural bioinformatics is a potent biological mechanism against phosphorous starvation in wheat. [ABSTRACT FROM AUTHOR]

Tumor necrosis factor α (TNF-α) is a multifunctional cytokine protein acknowledged as a vital mediator in cell differentiation, proliferation, and survival. Additionally, TNF-α is a crucial component of the host's defense by mediating inflammatory and immune responses against various aggressive agents, including viruses, bacteria parasites, and tumors. However, excessive production can be detrimental to the body and is also implicated in developing several inflammatory and immune-mediated disorders. Therefore, there is great interest in studying its role and its modulation, in various diseases, both in in vitro, in vivo, and in silico experiments. In this review, we evaluated the structures of proteins related to TNF-α available in public databases. In addition, we described the main antibodies blocking this cytokine and its applications and commented on the potential of naturally produced binding molecules, such as TNF-α-binding proteins produced by ticks. We also discuss the role of structural bioinformatics techniques in understanding the mechanisms of chronic inflammatory diseases related to TNF-α. We hope that the data presented in this review will be useful for studies that aim to better understand the mechanisms of the interactions of TNF-α with other proteins and will lead to new drugs or treatments. [ABSTRACT FROM AUTHOR]

Objective To express the protein enconded by the Rv3432c gene of Mycobacterium tuberculosis (M.tb) in vitro by prokaryotic expression, to analyze the structure of the Rv3432c protein by using bioinformatics software, and to explore for new drug targets against M.tb. Methods The Rv3432c gene was amplified by PCR using the genomic DNA of the inactivated M.tb strain H37Rv as the template and a recombinant plasmid was constructed with the expression vector pET-28a. The expression products were analyzed by SDS-PAGE and purified using affinity chromatography. The biological properties of Rv3432c were analyzed with Protparam, the Pfam online tool, SOMPA, Protscale, TMHMM Signalp 6.0, NetPhos3.1, SUMOsp 2.0, and SWISS-MODEL. Results pET-28a-Rv3432c recombinant plasmid sequencing results were fully consistent with those of the target gene. SDS-PAGE analysis showed that the fusion protein existed in the form of a soluble protein with a relative molecular mass of about 55×10³, which matched the expected size. ProtParam analysis showed that the Rv3432c protein was hydrophilic (showing a GRAVY value of -0.079). Rv3432c was a protein with no transmembrane structural domains or signal peptide. The secondary structure of Rv3432c mainly consisted of random coils (39.78%) and α-helix (39.57%) and was relatively loosely structured. Conclusion We successfully constructed a prokaryotic expression plasmid of the Rv3432c protein and analyzed its structure using bioinformatics, laying the foundation for further research on the role of Rv3432c in the pathogenesis and progression of tuberculosis as well as the identification of new drug targets against M.tb. [ABSTRACT FROM AUTHOR]

The prediction of three-dimensional (3D) protein structure from amino acid sequences has stood as a significant challenge in computational and structural bioinformatics for decades. Recently, the widespread integration of artificial intelligence (AI) algorithms has substantially expedited advancements in protein structure prediction, yielding numerous significant milestones. In particular, the end-to-end deep learning method AlphaFold2 has facilitated the rise of structure prediction performance to new heights, regularly competitive with experimental structures in the 14th Critical Assessment of Protein Structure Prediction (CASP14). To provide a comprehensive understanding and guide future research in the field of protein structure prediction for researchers, this review describes various methodologies, assessments, and databases in protein structure prediction, including traditionally used protein structure prediction methods, such as template-based modeling (TBM) and template-free modeling (FM) approaches; recently developed deep learning-based methods, such as contact/distance-guided methods, end-to-end folding methods, and protein language model (PLM)-based methods; multi-domain protein structure prediction methods; the CASP experiments and related assessments; and the recently released AlphaFold Protein Structure Database (AlphaFold DB). We discuss their advantages, disadvantages, and application scopes, aiming to provide researchers with insights through which to understand the limitations, contexts, and effective selections of protein structure prediction methods in protein-related fields. [ABSTRACT FROM AUTHOR]

N(1)-aminopropyl agmatine ureohydrolase (SpeB) is considered an essential enzyme for the growth and survival of thermophiles, it is involved in the biosynthesis of polyamines. The present study investigates the conformational stability and flexibility of Thermus thermophilus HB8 SpeB and its interactions with substrate, N1-aminopropyl agmatine at different physiological conditions to probe the optimal conditions that provide insights on biotechnological applications. As the 3D structure of SpeB is yet to be crystallized, it was modelled and validated using structural bioinformatics methods. To understand the conformational dynamics and also to assess the impact of pH and temperature variables on the tertiary structure of SpeB , atomistic molecular dynamics simulation (MDS) was employed. The MD investigation revealed that 300 K is not optimal for SpeB (apo form) at both acidic pH 4.5, and alkaline pH 8.5, since it exhibited decreased stability and compactness with higher residual flexibility. Further, the stable and strong binding of N1-aminopropyl agmatine with SpeB was observed at following conditions, neutral pH 7 at 353.15 K and alkaline pH 11 at 300 K. Thus, this in silico study provides significant insights into the structural investigations on SpeB along with optimal pH as well as temperature for its ideal enzymatic function. Highlights Structural and dynamic characteristics of T. thermophilus SpeB have been explored in detail Comparative Molecular dynamics simulations were performed to evaluate the effects of both pH and temperature on the structural stability of SpeB SpeB showed pH and temperature-dependent conformational changes that potentially affect substrate binding Structural stability of apo form of SpeB is not optimal at high temperature (363.15 K) in both acidic (pH 5.5) and neutral pH Stable and strong binding affinity of substrate molecule was observed in a neutral pH 7 at 353 K and alkaline pH 11 at 300K [ABSTRACT FROM AUTHOR]

Background: Machine learning (ML) has a rich history in structural bioinformatics, and modern approaches, such as deep learning, are revolutionizing our knowledge of the subtle relationships between biomolecular sequence, structure, function, dynamics and evolution. As with any advance that rests upon statistical learning approaches, the recent progress in biomolecular sciences is enabled by the availability of vast volumes of sufficiently-variable data. To be useful, such data must be well-structured, machine-readable, intelligible and manipulable. These and related requirements pose challenges that become especially acute at the computational scales typical in ML. Furthermore, in structural bioinformatics such data generally relate to protein three-dimensional (3D) structures, which are inherently more complex than sequence-based data. A significant and recurring challenge concerns the creation of large, high-quality, openly-accessible datasets that can be used for specific training and benchmarking tasks in ML pipelines for predictive modeling projects, along with reproducible splits for training and testing. Results: Here, we report 'Prop3D', a platform that allows for the creation, sharing and extensible reuse of libraries of protein domains, featurized with biophysical and evolutionary properties that can range from detailed, atomically-resolved physicochemical quantities (e.g., electrostatics) to coarser, residue-level features (e.g., phylogenetic conservation). As a community resource, we also supply a 'Prop3D-20sf' protein dataset, obtained by applying our approach to CATH. We have developed and deployed the Prop3D framework, both in the cloud and on local HPC resources, to systematically and reproducibly create comprehensive datasets via the Highly Scalable Data Service (HSDS). Our datasets are freely accessible via a public HSDS instance, or they can be used with accompanying Python wrappers for popular ML frameworks. Conclusion: Prop3D and its associated Prop3D-20sf dataset can be of broad utility in at least three ways. Firstly, the Prop3D workflow code can be customized and deployed on various cloud-based compute platforms, with scalability achieved largely by saving the results to distributed HDF5 files via HSDS. Secondly, the linked Prop3D-20sf dataset provides a hand-crafted, already-featurized dataset of protein domains for 20 highly-populated CATH families; importantly, provision of this pre-computed resource can aid the more efficient development (and reproducible deployment) of ML pipelines. Thirdly, Prop3D-20sf's construction explicitly takes into account (in creating datasets and data-splits) the enigma of 'data leakage', stemming from the evolutionary relationships between proteins. [ABSTRACT FROM AUTHOR]

Object retrieval systems measure the degree of similarity of the shape of 3D models. They search for the elements of the 3D model databases that resemble the query model. In structural bioinformatics, the query model is a protein tertiary/quaternary structure and the objective is to find similarly shaped molecules in the Protein Data Bank. With the ever-growing size of the PDB, a direct atomic coordinate comparison with all its members is impractical. To overcome this problem, the shape of the molecules can be encoded by fixed-length feature vectors. The distance of a protein to the entire PDB can be measured in this low-dimensional domain in linear time. The state-of-the-art approaches utilize Zernike–Canterakis moments for the shape encoding and supply the retrieval process with geometric data of the input structures. The BioZernike descriptors are a standard utility of the PDB since 2020. However, when trying to calculate the ZC moments locally, the issue of the deficiency of libraries readily available for use in custom programs (i.e., without relying on external binaries) is encountered, in particular programs written in Python. Here, a fast and well-documented Python implementation of the Pozo–Koehl algorithm is presented. In contrast to the more popular algorithm by Novotni and Klein, which is based on the voxelized volume, the PK algorithm produces ZC moments directly from the triangular surface meshes of 3D models. In particular, it can accept the molecular surfaces of proteins as its input. In the presented PK-Zernike library, owing to Numba's just-in-time compilation, a mesh with 50,000 facets is processed by a single thread in a second at the moment order 20. Since this is the first time the PK algorithm is used in structural bioinformatics, it is employed in a novel, simple, but efficient protein structure retrieval pipeline. The elimination of the outlying chain fragments via a fast PCA-based subroutine improves the discrimination ability, allowing for this pipeline to achieve an 0.961 area under the ROC curve in the BioZernike validation suite (0.997 for the assemblies). The correlation between the results of the proposed approach and of the 3D Surfer program attains values up to 0.99. [ABSTRACT FROM AUTHOR]

The article presents a framework for a Bioinformatics competition that focuses on 4 key aspects: structure, model, overview, and perspectives. Structure represents the organizational framework employed to coordinate the main tasks involved in the competition. Model showcases the competition design, which encompasses 3 phases. Overview presents our case study, the League of Brazilian Bioinformatics (LBB) 2nd Edition. Finally, the section on perspectives provides a brief discussion of the LBB 2nd Edition, along with insights and feedback from participants. LBB is a biannual team competition launched in 2019 to promote the ongoing training of human resources in Bioinformatics and Computational Biology in Brazil. LBB aims to stimulate ongoing training in Bioinformatics by encouraging participation in competitions, promoting the organization of future Bioinformatics competitions, and fostering the integration of the Bioinformatics and Computational Biology community in the country, as well as collaboration among participants. The LBB 2nd Edition was launched in 2021 and featured 251 competitors forming 91 teams. Knowledge competitions promote learning, collaboration, and innovation, which are crucial for advancing scientific knowledge and solving real-world problems. In summary, this article serves as a valuable resource for individuals and organizations interested in developing knowledge competitions, offering a model based on our experience with LBB to benefit all levels of Bioinformatics trainees. Author summary: There are various methods of learning, and one such approach involves a commonly used in the computing field, fostering competition, where communication, teamwork, group learning, and problem-solving play crucial roles. Similarly, in bioinformatics and computational biology, certain institutions or platforms promote learning through challenges. However, national-level competitions in these domains are relatively scarce. Here, we would like to present a framework for a Bioinformatics competition by showing structure, model, overview, and perspectives, using the League of Brazilian Bioinformatics (LBB) 2nd Edition as case study. LBB is the first bioinformatics competition in Latin America and aims to encompass the diverse multidisciplinary aspects of Bioinformatics and promote networking. In LBB, individuals not only have the opportunity to learn but also get to know one another. Further, to stimulate undergraduate training, we promoted teams with a mixture of different academic backgrounds. This approach promotes integration among participants and facilitates the exchange of information. To ensure that all participants have the opportunity to connect and form teams regardless of their background, we have created the MATCH, a script that allows individuals to be randomly paired with others, enabling them to get to know each other and create teams effectively. This approach fosters collaboration and creates a conducive environment for learning and problem-solving. [ABSTRACT FROM AUTHOR]

The program SSDraw generates publication‐quality protein secondary structure diagrams from three‐dimensional protein structures. To depict relationships between secondary structure and other protein features, diagrams can be colored by conservation score, B‐factor, or custom scoring. Diagrams of homologous proteins can be registered according to an input multiple sequence alignment. Linear visualization allows the user to stack registered diagrams, facilitating comparison of secondary structure and other properties among homologous proteins. SSDraw can be used to compare secondary structures of homologous proteins with both conserved and divergent folds. It can also generate one secondary structure diagram from an input protein structure of interest. The source code can be downloaded (https://github.com/ncbi/SSDraw) and run locally for rapid structure generation, while a Google Colab notebook allows easy use. [ABSTRACT FROM AUTHOR]

Ribonucleic acid (RNA) molecules serve as master regulators of cells by encoding their biological function in the ribonucleotide sequence, particularly their ability to interact with other molecules. To understand how RNA molecules perform their biological tasks and to design new sequences with specific functions, it is of great benefit to be able to computationally predict how RNA folds and interacts in the cellular environment. Our workflow for computational modeling of the 3D structures of RNA and its interactions with other molecules uses a set of methods developed in our laboratory, including MeSSPredRNA for predicting canonical and non-canonical base pairs, PARNASSUS for detecting remote homology based on comparisons of sequences and secondary structures, ModeRNA for comparative modeling, the SimRNA family of programs for modeling RNA 3D structure and its complexes with other molecules, and QRNAS for model refinement. In this study, we present the results of testing this workflow in predicting RNA 3D structures in the CASP15 experiment. The overall high score of the computational models predicted by our group demonstrates the robustness of our workflow and its individual components in terms of predicting RNA 3D structures of acceptable quality that are close to the target structures. However, the variance in prediction quality is still quite high, and the results are still too far from the level of protein 3D structure predictions. This exercise led us to consider several improvements, especially to better predict and enforce stacking interactions and non-canonical base pairs. [ABSTRACT FROM AUTHOR]

Aim: Glucose-regulated protein 78 (GRP78) is an endoplasmic reticulum located chaperone that plays a vital role during cellular stress. NS5A is one of the hepatitis C virus (HCV) non-structural proteins essential for replication. Materials & methods: Protein–protein docking was used to test the binding mode between GRP78 and HCV NS5A. Molecular dynamics simulations (MDSs) are performed on HCV NS5A, GRP78 and the HCV NS5A–GRP78 complex. Results: Docking and MDS reveal the ability of the GRP78 substrate-binding domain β to associate tightly with the HCV NS5A C142-C165 region. Conclusion: MDS reveals the potential of the C142-C165 region of the HCV NS5A to be used as a seed to develop a recognition inhibitor that counterparts the viral protein recognition by GRP78. Human cells contain a protein called GRP78. GRP78 is important to the ability of germs to cause infection because they can bind GRP78. We used a computer model to test how GRP78 binds to germs. This could be a target for future medicines and a useful way to stop infections. [ABSTRACT FROM AUTHOR]

The DSR-IBUN dextransucrase produced by Leuconostoc mesenteroides strain IBUN 91.2.98 has a short production time (4.5 hours), an enzymatic activity of 24.8 U/mL, and a specific activity of purified enzyme 2 times higher (331.6 U/mg) than that reported for similar enzymes. The aim of this study was to generate a structural model that, from an in silico approach, allows a better understanding, from the structural point of view, of the activity obtained by the enzyme of interest, which is key to continue with its study and industry application. For this, we translated the nucleotide sequence of the dsr_IBUN gene. With the primary structure of DSR-IBUN, the in silico prediction of physicochemical parameters, the possible subcellular localization, the presence of signal peptide, and the location of domains and functional and structural motifs of the protein were established. Subsequently, its secondary and tertiary structure were predicted and a homology model of the dextransucrase under study was constructed using Swiss-Model, performing careful template selection. The values obtained for the model, Global Model Quality Estimation (0.63), Quality Mean (−1.49), and root-mean-square deviation (0.09), allow us to affirm that the model for the enzyme dextransucrase DSR-IBUN is of adequate quality and can be used as a source of information for this protein. [ABSTRACT FROM AUTHOR]

In eukaryotic organisms, genomic DNA associates with histone proteins to form nucleosomes. Nucleosomes provide a basis for genome compaction, epigenetic markup, and mediate interactions of nuclear proteins with their target DNA loci. A negatively charged (acidic) patch located on the H2A-H2B histone dimer is a characteristic feature of the nucleosomal surface. The acidic patch is a common site in the attachment of various chromatin proteins, including viral ones. Acidic patch-binding peptides present perspective compounds that can be used to modulate chromatin functioning by disrupting interactions of nucleosomes with natural proteins or alternatively targeting artificial moieties to the nucleosomes, which may be beneficial for the development of new therapeutics. In this work, we used several computational and experimental techniques to improve our understanding of how peptides may bind to the acidic patch and what are the consequences of their binding. Through extensive analysis of the PDB database, histone sequence analysis, and molecular dynamic simulations, we elucidated common binding patterns and key interactions that stabilize peptide–nucleosome complexes. Through MD simulations and FRET measurements, we characterized changes in nucleosome dynamics conferred by peptide binding. Using fluorescence polarization and gel electrophoresis, we evaluated the affinity and specificity of the LANA1-22 peptide to DNA and nucleosomes. Taken together, our study provides new insights into the different patterns of intermolecular interactions that can be employed by natural and designed peptides to bind to nucleosomes, and the effects of peptide binding on nucleosome dynamics and stability. [ABSTRACT FROM AUTHOR]

This report describes the tertiary structure prediction assessment of difficult modeling targets in the 14th round of the Critical Assessment of Structure Prediction (CASP14). We implemented an official ranking scheme that used the same scores as the previous CASP topology-based assessment, but combined these scores with one that emphasized physically realistic models. The top performing AlphaFold2 group outperformed the rest of the prediction community on all but two of the difficult targets considered in this assessment. They provided high quality models for most of the targets (86% over GDT_TS 70), including larger targets above 150 residues, and they correctly predicted the topology of almost all the rest. AlphaFold2 performance was followed by two manual Baker methods, a Feig method that refined Zhang-server models, two notable automated Zhang server methods (QUARK and Zhang-server), and a Zhang manual group. Despite the remarkable progress in protein structure prediction of difficult targets, both the prediction community and AlphaFold2, to a lesser extent, faced challenges with flexible regions and obligate oligomeric assemblies. The official ranking of top-performing methods was supported by performance generated PCA and heatmap clusters that gave insight into target difficulties and the most successful state-of-the-art structure prediction methodologies.