High-confidence Models Research Articles

Prediction of conformational states in a coronavirus channel using Alphafold-2 and DeepMSA2: Strengths and limitations

The envelope (E) protein is present in all coronavirus genera. This protein can form pentameric oligomers with ion channel activity which have been proposed as a possible therapeutic target. However, high resolution structures of E channels are limited to those of the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), responsible for the recent COVID-19 pandemic. In the present work, we used Alphafold-2 (AF2), in ColabFold without templates, to predict the transmembrane domain (TMD) structure of six E-channels representative of genera alpha-, beta- and gamma-coronaviruses in the Coronaviridae family. High-confidence models were produced in all cases when combining multiple sequence alignments (MSAs) obtained from DeepMSA2. Overall, AF2 predicted at least two possible orientations of the α-helices in E-TMD channels: one where a conserved polar residue (Asn-15 in the SARS sequence) is oriented towards the center of the channel, ‘polar-in’, and one where this residue is in an interhelical orientation ‘polar-inter’. For the SARS models, the comparison with the two experimental models ‘closed’ (PDB: 7K3G) and ‘open’ (PDB: 8SUZ) is described, and suggests a ∼60˚ α-helix rotation mechanism involving either the full TMD or only its N-terminal half, to allow the passage of ions. While the results obtained are not identical to the two high resolution models available, they suggest various conformational states with striking similarities to those models. We believe these results can be further optimized by means of MSA subsampling, and guide future high resolution structural studies in these and other viral channels.

Benchmarking AlphaFold-Generated Structures of Chemokine-Chemokine Receptor Complexes.

AlphaFold and AlphaFold-Multimer have become two essential tools for the modeling of unknown structures of proteins and protein complexes. In this work, we extensively benchmarked the quality of chemokine-chemokine receptor structures generated by AlphaFold-Multimer against experimentally determined structures. Our analysis considered both the global quality of the model, as well as key structural features for chemokine recognition. To study the effects of template and multiple sequence alignment parameters on the results, a new prediction pipeline called LIT-AlphaFold (https://github.com/LIT-CCM-lab/LIT-AlphaFold) was developed, allowing extensive input customization. AlphaFold-Multimer correctly predicted differences in chemokine binding orientation and accurately reproduced the unique binding orientation of the CXCL12-ACKR3 complex. Further, the predictions of the full receptor N-terminus provided insights into a putative chemokine recognition site 0.5. The accuracy of chemokine N-terminus binding mode prediction varied between complexes, but the confidence score permitted the distinguishing of residues that were very likely well positioned. Finally, we generated a high-confidence model of the unsolved CXCL12-CXCR4 complex, which agreed with experimental mutagenesis and cross-linking data.

Gelato: a new hybrid deep learning-based Informer model for multivariate air pollution prediction.

The increase in air pollutants and its adverse effects on human health and the environment has raised significant concerns. This implies the necessity of predicting air pollutant levels. Numerous studies have aimed to provide new models for more accurate prediction of air pollutants such as CO2, O3, and PM2.5. Most of the models used in the literature are deep learning models with Transformers being the best for time series prediction. However, there is still a need to enhance accuracy in air pollution prediction using Transformers. Alongside the need for increased accuracy, there is a significant demand for predicting a broader spectrum of air pollutants. To encounter this challenge, this paper proposes a new hybrid deep learning-based Informer model called "Gelato" for multivariate air pollution prediction. Gelato takes a leap forward by taking several air pollutants into consideration simultaneously. Besides introducing new changes to the Informer structure as the base model, Gelato utilizes Particle Swarm Optimization for hyperparameter optimization. Moreover, XGBoost is used at the final stage to achieve minimal errors. Applying the proposed model on a dataset containing eight important air pollutants, including CO2, O3, NO, NO2, SO2, PM10, NH3, and PM2.5, the Gelato performance is assessed. Comparing the results of Gelato with other models shows Gelato's superiority over them, proving it is a high-confidence model for multivariate air pollution prediction.

Structure and function of SPP/SPPL proteases: insights from biochemical evidence and predictive modeling.

More than 20 years ago, signal peptide peptidase (SPP) and its homologues, the signal peptide peptidase-like (SPPL) proteases have been identified based on their sequence similarity to presenilins, a related family of intramembrane aspartyl proteases. Other than those for the presenilins, no high-resolution structures for the SPP/SPPL proteases are available. Despite this limitation, over the years bioinformatical and biochemical data have accumulated, which altogether have provided a picture of the overall structure and topology of these proteases, their localization in the cell, the process of substrate recognition, their cleavage mechanism, and their function. Recently, the artificial intelligence-based structure prediction tool AlphaFold has added high-confidence models of the expected fold of SPP/SPPL proteases. In this review, we summarize known structural aspects of the SPP/SPPL family as well as their substrates. Of particular interest are the emerging substrate recognition and catalytic mechanisms that might lead to the prediction and identification of more potential substrates and deeper insight into physiological and pathophysiological roles of proteolysis.

Predicted structural proteome of Sphagnum divinum and proteome-scale annotation.

Sphagnum-dominated peatlands store a substantial amount of terrestrial carbon. The genus is undersampled and under-studied. No experimental crystal structure from any Sphagnum species exists in the Protein Data Bank and fewer than 200 Sphagnum-related genes have structural models available in the AlphaFold Protein Structure Database. Tools and resources are needed to help bridge these gaps, and to enable the analysis of other structural proteomes now made possible by accurate structure prediction. We present the predicted structural proteome (25134 primary transcripts) of Sphagnum divinum computed using AlphaFold, structural alignment results of all high-confidence models against an annotated nonredundant crystallographic database of over 90,000 structures, a structure-based classification of putative Enzyme Commission (EC) numbers across this proteome, and the computational method to perform this proteome-scale structure-based annotation. All data and code are available in public repositories, detailed at https://github.com/BSDExabio/SAFA. The structural models of the S. divinum proteome have been deposited in the ModelArchive repository at https://modelarchive.org/doi/10.5452/ma-ornl-sphdiv.

The Trypanosoma brucei MISP family of invariant proteins is co-expressed with BARP as triple helical bundle structures on the surface of salivary gland forms, but is dispensable for parasite development within the tsetse vector.

Trypanosoma brucei spp. develop into mammalian-infectious metacyclic trypomastigotes inside tsetse salivary glands. Besides acquiring a variant surface glycoprotein (VSG) coat, little is known about the metacyclic expression of invariant surface antigens. Proteomic analyses of saliva from T. brucei-infected tsetse flies identified, in addition to VSG and Brucei Alanine-Rich Protein (BARP) peptides, a family of glycosylphosphatidylinositol (GPI)-anchored surface proteins herein named as Metacyclic Invariant Surface Proteins (MISP) because of its predominant expression on the surface of metacyclic trypomastigotes. The MISP family is encoded by five paralog genes with >80% protein identity, which are exclusively expressed by salivary gland stages of the parasite and peak in metacyclic stage, as shown by confocal microscopy and immuno-high resolution scanning electron microscopy. Crystallographic analysis of a MISP isoform (MISP360) and a high confidence model of BARP revealed a triple helical bundle architecture commonly found in other trypanosome surface proteins. Molecular modelling combined with live fluorescent microscopy suggests that MISP N-termini are potentially extended above the metacyclic VSG coat, and thus could be tested as a transmission-blocking vaccine target. However, vaccination with recombinant MISP360 isoform did not protect mice against a T. brucei infectious tsetse bite. Lastly, both CRISPR-Cas9-driven knock out and RNAi knock down of all MISP paralogues suggest they are not essential for parasite development in the tsetse vector. We suggest MISP may be relevant during trypanosome transmission or establishment in the vertebrate's skin.

Towards a structurally resolved human protein interaction network

Cellular functions are governed by molecular machines that assemble through protein-protein interactions. Their atomic details are critical to studying their molecular mechanisms. However, fewer than 5% of hundreds of thousands of human protein interactions have been structurally characterized. Here we test the potential and limitations of recent progress in deep-learning methods using AlphaFold2 to predict structures for 65,484 human protein interactions. We show that experiments can orthogonally confirm higher-confidence models. We identify 3,137 high-confidence models, of which 1,371 have no homology to a known structure. We identify interface residues harboring disease mutations, suggesting potential mechanisms for pathogenic variants. Groups of interface phosphorylation sites show patterns of co-regulation across conditions, suggestive of coordinated tuning of multiple protein interactions as signaling responses. Finally, we provide examples of how the predicted binary complexes can be used to build larger assemblies helping to expand our understanding of human cell biology.

Mapping the interaction between Trim28 and the KRAB domain at the center of Trim28 silencing of endogenous retroviruses.

Transcription of endogenous retroviral elements are tightly regulated during development by members of the KRAB‐containing zinc finger proteins (KRAB‐ZFPs) and the co‐repressor Trim28 (also known as Kap‐1 or Tif1β). KRAB‐ZFPs form the largest family of transcription regulators in mammals and initiate transcriptional silencing by tethering Trim28 to a target locus. Subsequently, Trim28 recruits chromatin modifying effectors resulting in the formation of heterochromatin. In the present study, we identify surface exposed residues on the central six turns of the Trim28 coiled‐coil region forming the binding interface for the KRAB domain. Using AlphaFold2 (AF2) we provide high confidence models of the interface between Trim28 and the KRAB domain and identified leucine 301 on each chain of the Trim28 monomer to act as a pin extending into a hydrophobic pocket on the KRAB domain surface. Site directed mutations in the Trim28‐KRAB binding interface abolished binding to the KRAB domain. Our work provides a detailed understanding of the specific interactions between the KRAB domain and the Trim28 coiled‐coil and how this interaction may be regulated during silencing events.

Characterization of Treponema denticola Major Surface Protein (Msp) by Deletion Analysis and Advanced Molecular Modeling.

Treponema denticola, a keystone pathogen in periodontitis, is a model organism for studying Treponema physiology and host-microbe interactions. Its major surface protein Msp forms an oligomeric outer membrane complex that binds fibronectin, has cytotoxic pore-forming activity, and disrupts several intracellular processes in host cells. T. denticola msp is an ortholog of the Treponema pallidum tprA to -K gene family that includes tprK, whose remarkable in vivo hypervariability is proposed to contribute to T. pallidum immune evasion. We recently identified the primary Msp surface-exposed epitope and proposed a model of the Msp protein as a β-barrel protein similar to Gram-negative bacterial porins. Here, we report fine-scale Msp mutagenesis demonstrating that both the N and C termini as well as the centrally located Msp surface epitope are required for native Msp oligomer expression. Removal of as few as three C-terminal amino acids abrogated Msp detection on the T. denticola cell surface, and deletion of four residues resulted in complete loss of detectable Msp. Substitution of a FLAG tag for either residues 6 to 13 of mature Msp or an 8-residue portion of the central Msp surface epitope resulted in expression of full-length Msp but absence of the oligomer, suggesting roles for both domains in oligomer formation. Consistent with previously reported Msp N-glycosylation, proteinase K treatment of intact cells released a 25 kDa polypeptide containing the Msp surface epitope into culture supernatants. Molecular modeling of Msp using novel metagenome-derived multiple sequence alignment (MSA) algorithms supports the hypothesis that Msp is a large-diameter, trimeric outer membrane porin-like protein whose potential transport substrate remains to be identified. IMPORTANCE The Treponema denticola gene encoding its major surface protein (Msp) is an ortholog of the T. pallidum tprA to -K gene family that includes tprK, whose remarkable in vivo hypervariability is proposed to contribute to T. pallidum immune evasion. Using a combined strategy of fine-scale mutagenesis and advanced predictive molecular modeling, we characterized the Msp protein and present a high-confidence model of its structure as an oligomer embedded in the outer membrane. This work adds to knowledge of Msp-like proteins in oral treponemes and may contribute to understanding the evolutionary and potential functional relationships between T. denticola Msp and the orthologous T. pallidum Tpr proteins.

AF2Complex predicts direct physical interactions in multimeric proteins with deep learning

Accurate descriptions of protein-protein interactions are essential for understanding biological systems. Remarkably accurate atomic structures have been recently computed for individual proteins by AlphaFold2 (AF2). Here, we demonstrate that the same neural network models from AF2 developed for single protein sequences can be adapted to predict the structures of multimeric protein complexes without retraining. In contrast to common approaches, our method, AF2Complex, does not require paired multiple sequence alignments. It achieves higher accuracy than some complex protein-protein docking strategies and provides a significant improvement over AF-Multimer, a development of AlphaFold for multimeric proteins. Moreover, we introduce metrics for predicting direct protein-protein interactions between arbitrary protein pairs and validate AF2Complex on some challenging benchmark sets and the E. coli proteome. Lastly, using the cytochrome c biogenesis system I as an example, we present high-confidence models of three sought-after assemblies formed by eight members of this system.

Adaptive response maps fusion of correlation filters with anti-occlusion mechanism for visual object tracking

Despite the impressive performance of correlation filter-based trackers in terms of robustness and accuracy, the trackers have room for improvement. The majority of existing trackers use a single feature or fixed fusion weights, which makes it possible for tracking to fail in the case of deformation or severe occlusion. In this paper, we propose a multi-feature response map adaptive fusion strategy based on the consistency of individual features and fused feature. It is able to improve the robustness and accuracy by building the better object appearance model. Moreover, since the response map has multiple local peaks when the target is occluded, we propose an anti-occlusion mechanism. Specifically, if the nonmaximal local peak is satisfied with our proposed conditions, we generate a new response map which is obtained by moving the center of the region of interest to the nonmaximal local peak position of the response map and re-extracting features. We then select the response map with the largest response value as the final response map. This proposed anti-occlusion mechanism can effectively cope with the problem of tracking failure caused by occlusion. Finally, by adjusting the learning rate in different scenes, we designed a high-confidence model update strategy to deal with the problem of model pollution. Besides, we conducted experiments on OTB2013, OTB2015, TC128 and UAV123 datasets and compared them with the current state-of-the-art algorithms, and the proposed algorithms have impressive advantages in terms of accuracy and robustness.

What can we learn from the relationship between gait and clinical impairments in children with cerebral palsy when comparing databases?

Spastic cerebral palsy (CP) is characterized by increased joint resistance, caused by a mix of increased tissue stiffness, as well as involuntary reflex and background muscle activity. These properties can be quantified using a neuromechanical model of the musculoskeletal complex and instrumented assessment. The construct validity of the neuromechanical parameters was examined (i.e. the internal model validity, effect of knee angle, speed and age, sensitivity to patients versus controls, spasticity severity and treatment), together with the repeatability. We included 38 children with CP and 35 controls. A motor driven footplate applied two slow (15°/s) and two fast (100°/s) rotations around the ankle joint, at two different knee angles. Ankle angle, torque and EMG of the gastrocnemius (GA), soleus (SO) and tibialis anterior (TA) muscle were used to optimize a nonlinear neuromuscular model. Outcome measures were tissue stiffness, reflex and background activity for GA, SO and TA. The internal model validity showed medium to high parameter confidence and good model fits. All parameter could discriminate between patients with CP and controls according to CP pathology. Other measures of external model validity (effect of test position, speed and age) showed behaviour along the lines of current knowledge of physiology. GA/SO background activity was sensitive to spasticity severity, but reflex activity was not. Preliminary data indicated that reflex activity was reduced after spasticity treatment. The between-trial and -day repeatability was moderate to good. The large variance between patients in the ratio of stiffness and neural resistance indicates that the method could potentially contribute to patient-specific treatment selection.

Epitope and Paratope Mapping of PD-1/Nivolumab by Mass Spectrometry-Based Hydrogen–Deuterium Exchange, Cross-linking, and Molecular Docking

Programmed cell death-1 (PD-1), an antigen co-receptor on cell surfaces, is one of the conspicuous immune checkpoints. Nivolumab, a monoclonal antibody therapeutic approved by the FDA, binds to PD-1 and efficiently blocks its pathways. In this study, an integrated approach was developed to map the epitope/paratope of PD-1/nivolumab. The approach includes hydrogen-deuterium exchange mass spectrometry (HDX-MS) followed by electron-transfer dissociation (ETD), chemical cross-linking, and molecular docking. HDX-ETD offers some binding-site characterization with amino acid resolution. Chemical cross-linking provides complementary information on one additional epitope (i.e., the BC-loop) and a potential paratope at the N-terminus of the heavy chain. Furthermore, cross-linking identifies another loop region (i.e., the C'D-loop) that undergoes a remote conformational change. The distance restraints derived from the cross-links enable building high-confidence models of PD-1/nivolumab, evaluated with respect to a resolved crystal structure. This integrated strategy is an opportunity to characterize comprehensively other antigen-antibody interactions, to enable the understanding of binding mechanisms, and to design future antibody therapeutics.

End-to-end DeepNCC framework for robust visual tracking

In this paper, we propose an NCC-based object tracking deep framework, which can be well initialized with the limited target samples in the first frame. The proposed framework contains a pretrained model, online feature fine-tuning layers and tracking processes. The pretrained model provides rich feature representations while online feature fine-tuning layers select discriminative and generic features for the tracked object. We choose normalized cross-correlation as a template tracking layer to perform the tracking process. To enable the learned features representation closely coordinated to the tracked target, we jointly train the feature representation network and tracking processes. In online tracking, an adaptive template and a fixed template are fused to find the optimal tracking results. Scale estimation and a high-confidence model update scheme are perfectly integrated into the framework to adapt to the target appearance changes. The extensive experiments demonstrate that the proposed tracker achieves superior performance compared with other state-of-the-art trackers.

Vibration Fracture Mechanism and Optimization Design of Array Power Supply for Near-Space SAR

The random vibration failure of an array power supply for near-space SAR was analyzed. The fracture mechanism and the fracture reason of fracture formation in the specimen were investigated. The results show that antishock MOS pin breaks first, and the power supply is still in the working state during the process of random vibration. This caused dischargings at the tip of the fracture and melting of the tip of the broken pin which form a river-shaped fracture and granular tissue. The plastic fracture with typical dimple morphology of the pins for the resistor tube occurred during the random vibration. The intergranular fracture appeared at the welding part of the electronic components for array power supply, which presented a brittle fracture mechanism. The fracture was dominated by a ductile fracture for components when the stress produced by the vibration was close to the yield strength of the material. The fracture was dominated by a brittle fracture for components when the stress produced by the vibration was far beyond the yield strength of the material. A simulation evaluation system based on the high-confidence model was proposed. The stress of the electronic components for array power supply and its welding was much lower than the allowable strength of the material by the optimization of the structure and the form of the welding for the array power supply. The sample was successfully tested and verified without any further fracture problems.

Confidence-Constrained Support Vector Regression for Geological Surface Uncertainty Modeling

Reconstruction of complex geological surface is widely used in oil and gas exploration, geological modeling, geological structure analysis, and other fields. It is an important basis for data visualization and visual analysis in these fields. The complexity of geological structures, the inaccuracy and sparsity of seismic interpretation data, and the lack of tectonic morphological information can lead to uncertainty in geological surface reconstruction. The existing geological surface uncertainty characterization and uncertain reconstruction methods have a shortcoming in balancing the interpolation error of high-confidence samples and model structure risk. Based on support vector regression (SVR), a method with confidence constraints for uncertainty characterization and the modeling of geological surfaces is proposed in this article. The proposed method minimizes the structural risk by adding a regularization term representing the model complexity, integrates high-confidence samples, such as drilling data, based on confidence constraints, and utilizes well path points by assigning appropriate inequality constraints to the corresponding prediction points. The results based on a real-world fault data set show that the uncertainty envelopes and fault realizations generated by the proposed method are constrained by well observations and well paths, effectively reducing the uncertainty.

Connected and Shared X-in-the-Loop Technologies for Electric Vehicle Design

The presented paper introduces a new methodology of experimental testing procedures required by the complex systems of electric vehicles (EV). This methodology is based on real-time connection of test setups and platforms, which may be situated in different geographical locations, belong to various cyber-physical domains, and are united in a global X-in-the-loop (XIL) experimental environment. The proposed concept, called XILforEV, allows exploring interdependencies between various physical processes that can be identified or investigated in the process of EV development. The paper discusses the following relevant topics: global XILforEV architecture; realization of required high-confidence models using dynamic data driven application systems (DDDAS) and multi fidelity models (MFM) approaches; and formulation of case studies to illustrate XILforEV applications.

A high-confidence model updating correlation filtering tracker with scale adaptation for visual target tracking

In recent years, correlation filters have achieved excellent performance in terms of accuracy and robustness for visual target tracking. However, it is still a challenging task to improve the tracking capability which can handle complex background and deal with target scale changes. In this paper, a new tracker based on the correlation filter framework with scale adaptation for visual target tracking is presented: 1) an effective scale adaptive scheme, which can solve the fixed template size with the basis of traditional correlation filter tracker, is proposed; 2) two powerful features including histogram of oriented gradient and color name, which can improve the overall tracking performance, are integrated together; 3) a high-confidence model updating strategy, which can avoid the model occlusion problem during model updating, is developed. The extensive experimental evaluations on OTB-2013 dataset have demonstrated that the developed novel tracker in our study can successfully outperform the other trackers on the benchmark dataset with 51 sequences. It is very promising for various challenging scenarios.

On the Prediction of Upwelling Events at the Colombian Caribbean Coasts from Modis-SST Imagery.

The upwelling cores on the Caribbean Colombian coasts are mainly located at the Peninsula de la Guajira and Cabo de la Aguja. We used monthly averaged Moderate Resolution Imaging Spectroradiometer (MODIS) sea surface temperature as the only information to build up a prediction model for the upwelling events. This comprised two steps: (i) the reduction of the complexity by means of the Karhunen–Loève transform and (ii) a prediction model of time series. Two prediction models were considered: (a) a parametric autoregressive-moving average (ARMA) time series from the Box–Jenkins methodology and (b) a harmonic synthesis model. The harmonic synthesis also comprised of two steps: the maximum entropy spectral analysis and a least-squares harmonic analysis on the set of frequencies. The parametric ARMA time series model failed at the time of prediction with a very narrow range, and it was quite difficult to apply. The harmonic synthesis allowed prediction with a horizon of six months with a correlation of about 0.80. The results can be summarized using the time series of the weights of the different oscillation modes, their spatial structures with the nodal lines, and a high confidence model with a horizon of prediction of about four months.

Annotation of the Giardia proteome through structure-based homology and machine learning.

BackgroundLarge-scale computational prediction of protein structures represents a cost-effective alternative to empirical structure determination with particular promise for non-model organisms and neglected pathogens. Conventional sequence-based tools are insufficient to annotate the genomes of such divergent biological systems. Conversely, protein structure tolerates substantial variation in primary amino acid sequence and is thus a robust indicator of biochemical function. Structural proteomics is poised to become a standard part of pathogen genomics research; however, informatic methods are now required to assign confidence in large volumes of predicted structures.AimsOur aim was to predict the proteome of a neglected human pathogen, Giardia duodenalis, and stratify predicted structures into high- and lower-confidence categories using a variety of metrics in isolation and combination.MethodsWe used the I-TASSER suite to predict structural models for ∼5,000 proteins encoded in G. duodenalis and identify their closest empirically-determined structural homologues in the Protein Data Bank. Models were assigned to high- or lower-confidence categories depending on the presence of matching protein family (Pfam) domains in query and reference peptides. Metrics output from the suite and derived metrics were assessed for their ability to predict the high-confidence category individually, and in combination through development of a random forest classifier.ResultsWe identified 1,095 high-confidence models including 212 hypothetical proteins. Amino acid identity between query and reference peptides was the greatest individual predictor of high-confidence status; however, the random forest classifier outperformed any metric in isolation (area under the receiver operating characteristic curve = 0.976) and identified a subset of 305 high-confidence-like models, corresponding to false-positive predictions. High-confidence models exhibited greater transcriptional abundance, and the classifier generalized across species, indicating the broad utility of this approach for automatically stratifying predicted structures. Additional structure-based clustering was used to cross-check confidence predictions in an expanded family of Nek kinases. Several high-confidence-like proteins yielded substantial new insight into mechanisms of redox balance in G. duodenalis—a system central to the efficacy of limited anti-giardial drugs.ConclusionStructural proteomics combined with machine learning can aid genome annotation for genetically divergent organisms, including human pathogens, and stratify predicted structures to promote efficient allocation of limited resources for experimental investigation.