Neural Upscaling from Residue-Level Protein Structure Networks to Atomistic Structures
Explore the source record for details and available documents.
Vy T. Duong↗
SEARCH · Engineering Papers
Results for “Protein structure network”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Protein conformational switch discerned via network centrality properties
Network analysis has emerged as a powerful tool for examining structural biology systems. The spatial organization of the components of a biomolecular structure has been rendered as a graph representation and analyses have been performed to deduce the biophysical and mechanistic properties of these components. For proteins, the analysis of protein structure networks (PSNs), especially via network centrality measurements and cluster coefficients, has led to identifying amino acid residues that play key functional roles and classifying amino acid residues in general. Whether these network properties examined in various studies are sensitive to subtle (yet biologically significant) conformational changes remained to be addressed. Here, we focused on four types of network centrality properties (betweenness, closeness, degree, and eigenvector centralities) for conformational changes upon ligand binding of a sensor protein (constitutive androstane receptor) and an allosteric enzyme (ribonucleotide reductase). We found that eigenvector centrality is sensitive and can distinguish salient structural features between protein conformational states while other centrality measures, especially closeness centrality, are less sensitive and rather generic with respect to the structural specificity. We also demonstrated that an ensembleinformed, modified PSN with static edges removed (which we term PSN*) has enhanced sensitivity at discerning structural changes.
59 BASIC BIOLOGICAL SCIENCES↗
Allosteric prediction via convolutional neural networks and protein structural and dynamical features
Allostery is the phenomenon whereby a binding event or covalent modification at one site in a protein modulates function at a distal site, thus changing a protein’s functional state. As such, it is a ubiquitous aspect of protein functional regulation. Computationally predicting allosteric states is important as part of the broader challenge of functional annotation, but it also has practical implications for drug development, as targeting an allosteric site often affords greater specificity compared with targeting an orthosteric site. This study introduces a machine learning approach to predict the allosteric functional state using the small G-protein KRas as the model system, due to its implication in many types of cancer and being well studied as a result with many x-ray crystallographic structures of KRas available with different mutations and ligands bound. Using structural and dynamical features that can be cast as images, namely interatomic distances, contact maps, covariance, and mutual information, supervised learning was performed using convolutional neural networks. Two pretrained convolutional neural network architectures, GoogLeNet and ResNet18, were fine-tuned to classify KRas into active or inactive states based on these features. Across training regimes, atomic contact maps emerged as the most effective structural feature, whereas linearized mutual information outperformed covariance in capturing dynamical correlations relevant to allostery. Models achieved significant validation accuracy, with atomic contact maps yielding up to 90% accuracy. In conclusion, the findings suggest that integrating global structural rearrangements and correlated motion patterns with deep learning can reliably predict protein allosteric states, offering a promising framework for understanding allosteric regulation and developing targeted therapeutics.
Rajeshwar T., Rajitha [Oak Ridge National Laborato↗
De novo design of protein structure and function with RFdiffusion
Abstract There has been considerable recent progress in designing new proteins using deep-learning methods 1–9 . Despite this progress, a general deep-learning framework for protein design that enables solution of a wide range of design challenges, including de novo binder design and design of higher-order symmetric architectures, has yet to be described. Diffusion models 10,11 have had considerable success in image and language generative modelling but limited success when applied to protein modelling, probably due to the complexity of protein backbone geometry and sequence–structure relationships. Here we show that by fine-tuning the RoseTTAFold structure prediction network on protein structure denoising tasks, we obtain a generative model of protein backbones that achieves outstanding performance on unconditional and topology-constrained protein monomer design, protein binder design, symmetric oligomer design, enzyme active site scaffolding and symmetric motif scaffolding for therapeutic and metal-binding protein design. We demonstrate the power and generality of the method, called RoseTTAFold diffusion (RFdiffusion), by experimentally characterizing the structures and functions of hundreds of designed symmetric assemblies, metal-binding proteins and protein binders. The accuracy of RFdiffusion is confirmed by the cryogenic electron microscopy structure of a designed binder in complex with influenza haemagglutinin that is nearly identical to the design model. In a manner analogous to networks that produce images from user-specified inputs, RFdiffusion enables the design of diverse functional proteins from simple molecular specifications.
Science & Technology - Other Topics↗
Combining protein sequences and structures with transformers and equivariant graph neural networks to predict protein function
Abstract Motivation Millions of protein sequences have been generated by numerous genome and transcriptome sequencing projects. However, experimentally determining the function of the proteins is still a time consuming, low-throughput, and expensive process, leading to a large protein sequence-function gap. Therefore, it is important to develop computational methods to accurately predict protein function to fill the gap. Even though many methods have been developed to use protein sequences as input to predict function, much fewer methods leverage protein structures in protein function prediction because there was lack of accurate protein structures for most proteins until recently. Results We developed TransFun—a method using a transformer-based protein language model and 3D-equivariant graph neural networks to distill information from both protein sequences and structures to predict protein function. It extracts feature embeddings from protein sequences using a pre-trained protein language model (ESM) via transfer learning and combines them with 3D structures of proteins predicted by AlphaFold2 through equivariant graph neural networks. Benchmarked on the CAFA3 test dataset and a new test dataset, TransFun outperforms several state-of-the-art methods, indicating that the language model and 3D-equivariant graph neural networks are effective methods to leverage protein sequences and structures to improve protein function prediction. Combining TransFun predictions and sequence similarity-based predictions can further increase prediction accuracy. Availability and implementation The source code of TransFun is available at https://github.com/jianlin-cheng/TransFun.
59 BASIC BIOLOGICAL SCIENCES↗
Comparative genomic analysis of thermophilic fungi reveals convergent evolutionary adaptations and gene losses
Thermophily is a trait scattered across the fungal tree of life, with its highest prevalence within three fungal families (Chaetomiaceae, Thermoascaceae, and Trichocomaceae), as well as some members of the phylum Mucoromycota. We examined 37 thermophilic and thermotolerant species and 42 mesophilic species for this study and identified thermophily as the ancestral state of all three prominent families of thermophilic fungi. Thermophilic fungal genomes were found to encode various thermostable enzymes, including carbohydrate-active enzymes such as endoxylanases, which are useful for many industrial applications. At the same time, the overall gene counts, especially in gene families responsible for microbial defense such as secondary metabolism, are reduced in thermophiles compared to mesophiles. We also found a reduction in the core genome size of thermophiles in both the Chaetomiaceae family and the Eurotiomycetes class. The Gene Ontology terms lost in thermophilic fungi include primary metabolism, transporters, UV response, and O-methyltransferases. Comparative genomics analysis also revealed higher GC content in the third base of codons (GC3) and a lower effective number of codons in fungal thermophiles than in both thermotolerant and mesophilic fungi. Furthermore, using the Support Vector Machine classifier, we identified several Pfam domains capable of discriminating between genomes of thermophiles and mesophiles with 94% accuracy. Using AlphaFold2 to predict protein structures of endoxylanases (GH10), we built a similarity network based on the structures. We found that the number of disulfide bonds appears important for protein structure, and the network clusters based on protein structures correlate with the optimal activity temperature. Thus, comparative genomics offers new insights into the biology, adaptation, and evolutionary history of thermophilic fungi while providing a parts list for bioengineering applications.
59 BASIC BIOLOGICAL SCIENCES↗
Binding Affinity Prediction by Pairwise Function Based on Neural Network
In this paper, we present a new approach to estimate the binding affinity from given three-dimensional poses of protein–ligand complexes. In this scheme, every protein–ligand atom pair makes an additive free-energy contribution. The sum of these pairwise contributions then gives the total binding free energy or the logarithm of the dissociation constant. The pairwise contribution is calculated by a function implemented via a neural network that takes the properties of the two atoms and their distance as input. The pairwise function is trained using a portion of the PDBbind 2018 data set. The model achieves good accuracy for affinity predictions when evaluated with PDBbind 2018 and with the CASF-2016 benchmark, comparing favorably to many scoring functions such as that of AutoDock Vina. The framework here may be extended to incorporate other factors to further improve its accuracy and power.
59 BASIC BIOLOGICAL SCIENCES↗
Hierarchical, rotation‐equivariant neural networks to select structural models of protein complexes
Abstract Predicting the structure of multi‐protein complexes is a grand challenge in biochemistry, with major implications for basic science and drug discovery. Computational structure prediction methods generally leverage predefined structural features to distinguish accurate structural models from less accurate ones. This raises the question of whether it is possible to learn characteristics of accurate models directly from atomic coordinates of protein complexes, with no prior assumptions. Here we introduce a machine learning method that learns directly from the 3D positions of all atoms to identify accurate models of protein complexes, without using any precomputed physics‐inspired or statistical terms. Our neural network architecture combines multiple ingredients that together enable end‐to‐end learning from molecular structures containing tens of thousands of atoms: a point‐based representation of atoms, equivariance with respect to rotation and translation, local convolutions, and hierarchical subsampling operations. When used in combination with previously developed scoring functions, our network substantially improves the identification of accurate structural models among a large set of possible models. Our network can also be used to predict the accuracy of a given structural model in absolute terms. The architecture we present is readily applicable to other tasks involving learning on 3D structures of large atomic systems.
Eismann, Stephan↗
High throughput, accurate gene annotation through AI and HPC-enabled structural analysis
With the advances in next generation sequencing technologies, the number of sequenced genomes is growing exponentially, resulting in a technology bottleneck for the translation of sequence information into usable hypotheses about the function of each gene. We have proposed leveraging our leadership high-performance computing (HPC) resources to help break this annotation bottleneck. Here we design an HPC-based framework to infer gene function from gene sequence by incorporating information about protein structure and interactions predicted by deep learning approaches. Accurate functional prediction and gene annotation using computational methods will facilitate breakthroughs in the genomic sciences essential to understanding and harnessing life processes in bacteria, fungi and plants. The development and applications of the state-of-the-art deep neural networks to protein structural modeling, interaction prediction, sequence comparison, and quality assessment of protein structural models will be made possible by leadership computational resources. These HPC-enabled bioinformatics and molecular modeling tools will provide powerful insights into molecular functions of genes.
59 BASIC BIOLOGICAL SCIENCES↗
Accurate prediction of protein structures and interactions using a three-track neural network
DeepMind presented notably accurate predictions at the recent 14th Critical Assessment of Structure Prediction (CASP14) conference. We explored network architectures that incorporate related ideas and obtained the best performance with a three-track network in which information at the one-dimensional (1D) sequence level, the 2D distance map level, and the 3D coordinate level is successively transformed and integrated. The three-track network produces structure predictions with accuracies approaching those of DeepMind in CASP14, enables the rapid solution of challenging x-ray crystallography and cryo–electron microscopy structure modeling problems, and provides insights into the functions of proteins of currently unknown structure. The network also enables rapid generation of accurate protein-protein complex models from sequence information alone, short-circuiting traditional approaches that require modeling of individual subunits followed by docking. Here, we make the method available to the scientific community to speed biological research.
59 BASIC BIOLOGICAL SCIENCES↗
Construction of extracellular matrix-based 3D hydrogel and its effects on cardiomyocytes
Highlights: • SOD and Triton X-100 could efficiently decellularize cells to get cardiac ECM. • Main proteins of cECM were fibronectin, collagen, elastin, fibrillin, laminin, etc. • 75% cECM-containing 3D hydrogel promoted cell proliferation and anti-apoptosis. • cECM3DH facilitated the differentiation of ESCs into CMs. • APcECM3DH was more conducive to NCM and ESC-derived CM maturation. Some discoveries resulted from 2-dimensional (2D) cultured cardiac cells have been disqualified in animal testing and later clinical trials. Extracellular matrix (ECM) plays a vital role in cardiac homeostasis, cardiac ECM (cECM)-based 3D cell cultures can mimics the physiological and pathological conditions in vivo closely, it is hopeful of addressing this challenge. Construction of cECM-based 3-dimensional (3D) hydrogel (cECM3DH) and its effects on cell behaviors were studied here. The results indicated that cellular compartments could be efficiently removed from heart tissue via sodium dodecyl sulfonate (SDS)- and Triton X-100-mediated decellularization, remaining the natural fibrous network structure and major proteins. 3D hydrogel consisted of 1 × 10{sup 7} cells/mL cells and 75% cECM could promote the proliferation and anti-apoptosis ability of human embryonic kidney (HEK)-293T cells. 0.25% trypsin or 0.20% collagenase was suitable to retrieve these cells from 3D hydrogel for further researches. Compared with 2D culture system, cECM3DH could significantly increase the proportion of GATA 4{sup +} cardiomyocytes (CMs) derived from heart tissue of neonatal mouse or induced differentiation of embryonic stem cells (ESCs) (P < 0.05) The expression levels of mature genes including cTnT, JCN, CaV1.2, MYL2, CASQ2, NCX1, and Cx43 of these CMs in adult pig cECM-based 3D hydrogel (APcECM3DH) were significantly higher than that in 2D culture system and in newborn piglet cECM-based 3D hydrogel (NPcECM3DH), respectively (P < 0.05). Therefore, cECM3DH supports the generation of primary CMs and ESC-derived CMs, APcECM3DH was more conducive to promoting CM maturation, which contributes to building 3D model for pathogenesis exploration, drug screening, and regenerative medicine of heart diseases.
60 APPLIED LIFE SCIENCES↗
Graph identification of proteins in tomograms ( GRIP‐Tomo )
Abstract In this study, we present a method of pattern mining based on network theory that enables the identification of protein structures or complexes from synthetic volume densities, without the knowledge of predefined templates or human biases for refinement. We hypothesized that the topological connectivity of protein structures is invariant, and they are distinctive for the purpose of protein identification from distorted data presented in volume densities. Three‐dimensional densities of a protein or a complex from simulated tomographic volumes were transformed into mathematical graphs as observables. We systematically introduced data distortion or defects such as missing fullness of data, the tumbling effect, and the missing wedge effect into the simulated volumes, and varied the distance cutoffs in pixels to capture the varying connectivity between the density cluster centroids in the presence of defects. A similarity score between the graphs from the simulated volumes and the graphs transformed from the physical protein structures in point data was calculated by comparing their network theory order parameters including node degrees, betweenness centrality, and graph densities. By capturing the essential topological features defining the heterogeneous morphologies of a network, we were able to accurately identify proteins and homo‐multimeric complexes from 10 topologically distinctive samples without realistic noise added. Our approach empowers future developments of tomogram processing by providing pattern mining with interpretability, to enable the classification of single‐domain protein native topologies as well as distinct single‐domain proteins from multimeric complexes within noisy volumes.
59 BASIC BIOLOGICAL SCIENCES↗
Solvent organization in the ultrahigh-resolution crystal structure of crambin at room temperature
Ultrahigh-resolution structures provide unprecedented details about protein dynamics, hydrogen bonding and solvent networks. The reported 0.70 Å, room-temperature crystal structure of crambin is the highest-resolution ambient-temperature structure of a protein achieved to date. Sufficient data were collected to enable unrestrained refinement of the protein and associated solvent networks using SHELXL . Dynamic solvent networks resulting from alternative side-chain conformations and shifts in water positions are revealed, demonstrating that polypeptide flexibility and formation of clathrate-type structures at hydrophobic surfaces are the key features endowing crambin crystals with extraordinary diffraction power.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
3D-equivariant graph neural networks for protein model quality assessment
Quality assessment (QA) of predicted protein tertiary structure models plays an important role in ranking and using them. With the recent development of deep learning end-to-end protein structure prediction techniques for generating highly confident tertiary structures for most proteins, it is important to explore corresponding QA strategies to evaluate and select the structural models predicted by them since these models have better quality and different properties than the models predicted by traditional tertiary structure prediction methods. We develop EnQA, a novel graph-based 3D-equivariant neural network method that is equivariant to rotation and translation of 3D objects to estimate the accuracy of protein structural models by leveraging the structural features acquired from the state-of-the-art tertiary structure prediction method—AlphaFold2. We train and test the method on both traditional model datasets (e.g. the datasets of the Critical Assessment of Techniques for Protein Structure Prediction) and a new dataset of high-quality structural models predicted only by AlphaFold2 for the proteins whose experimental structures were released recently. Our approach achieves state-of-the-art performance on protein structural models predicted by both traditional protein structure prediction methods and the latest end-to-end deep learning method—AlphaFold2. It performs even better than the model QA scores provided by AlphaFold2 itself. The results illustrate that the 3D-equivariant graph neural network is a promising approach to the evaluation of protein structural models. Integrating AlphaFold2 features with other complementary sequence and structural features is important for improving protein model QA.
59 BASIC BIOLOGICAL SCIENCES↗
DISTEMA: distance map-based estimation of single protein model accuracy with attentive 2D convolutional neural network
Abstract Background Estimation of the accuracy (quality) of protein structural models is important for both prediction and use of protein structural models. Deep learning methods have been used to integrate protein structure features to predict the quality of protein models. Inter-residue distances are key information for predicting protein’s tertiary structures and therefore have good potentials to predict the quality of protein structural models. However, few methods have been developed to fully take advantage of predicted inter-residue distance maps to estimate the accuracy of a single protein structural model. Result We developed an attentive 2D convolutional neural network (CNN) with channel-wise attention to take only a raw difference map between the inter-residue distance map calculated from a single protein model and the distance map predicted from the protein sequence as input to predict the quality of the model. The network comprises multiple convolutional layers, batch normalization layers, dense layers, and Squeeze-and-Excitation blocks with attention to automatically extract features relevant to protein model quality from the raw input without using any expert-curated features. We evaluated DISTEMA’s capability of selecting the best models for CASP13 targets in terms of ranking loss of GDT-TS score. The ranking loss of DISTEMA is 0.079, lower than several state-of-the-art single-model quality assessment methods. Conclusion This work demonstrates that using raw inter-residue distance information with deep learning can predict the quality of protein structural models reasonably well. DISTEMA is freely at https://github.com/jianlin-cheng/DISTEMA
59 BASIC BIOLOGICAL SCIENCES↗
Graph-Directed Approach for Downselecting Toxins for Experimental Structure Determination
Conotoxins are short, cysteine-rich peptides of great interest as novel therapeutic leads and of great concern as lethal biological agents due to their high affinity and specificity for various receptors involved in neuromuscular transmission. Currently, of the approximately 6000 known conotoxin sequences, only about 3% have associated structural characterization, which leads to a bottleneck in rapid high-throughput screening (HTS) for identification of potential leads or threats. In this work, we combine a graph-based approach with homology modeling to expand the library of conotoxin structures and to identify those conotoxin sequences that are of the greatest value for experimental structural characterization. The latter would allow for the rapid expansion of the known structural space for generating high quality template-based models. Our approach generalizes to other evolutionarily-related, short, cysteine-rich venoms of interest. Overall, we present and validate an approach for venom structure modeling and experimental guidance and employ it to produce a 290%-larger library of approximate conotoxin structures for HTS. We also provide a set of ranked conotoxin sequences for experimental structure determination to further expand this library.
59 BASIC BIOLOGICAL SCIENCES↗
Network analysis uncovers the communication structure of SARS-CoV-2 spike protein identifying sites for immunogen design
The COVID-19 pandemic, caused by the SARS-CoV-2 virus, has triggered myriad efforts to understand the structure and dynamics of this complex pathogen. The spike glycoprotein of SARS-CoV-2 is a significant target for immunogens as it is the means by which the virus enters human cells, while simultaneously sporting mutations responsible for immune escape. These functional and escape processes are regulated by complex molecular-level interactions. Our study presents quantitative insights on domain and residue contributions to allosteric communication, immune evasion, and local- and global-level control of functions through the derivation of a weighted graph representation from all-atom MD simulations. Focusing on the ancestral form and the D614G-variant, we provide evidence of the utility of our approach by guiding the selection of a mutation that alters the spike’s stability. Taken together, the network approach serves as a valuable tool to evaluate communication “hot-spots” in proteins to guide design of stable immunogens.
60 APPLIED LIFE SCIENCES↗
Atomic protein structure refinement using all-atom graph representations and SE(3)-equivariant graph transformer
Abstract Motivation The state-of-art protein structure prediction methods such as AlphaFold are being widely used to predict structures of uncharacterized proteins in biomedical research. There is a significant need to further improve the quality and nativeness of the predicted structures to enhance their usability. In this work, we develop ATOMRefine, a deep learning-based, end-to-end, all-atom protein structural model refinement method. It uses a SE(3)-equivariant graph transformer network to directly refine protein atomic coordinates in a predicted tertiary structure represented as a molecular graph. Results The method is first trained and tested on the structural models in AlphaFoldDB whose experimental structures are known, and then blindly tested on 69 CASP14 regular targets and 7 CASP14 refinement targets. ATOMRefine improves the quality of both backbone atoms and all-atom conformation of the initial structural models generated by AlphaFold. It also performs better than two state-of-the-art refinement methods in multiple evaluation metrics including an all-atom model quality score—the MolProbity score based on the analysis of all-atom contacts, bond length, atom clashes, torsion angles, and side-chain rotamers. As ATOMRefine can refine a protein structure quickly, it provides a viable, fast solution for improving protein geometry and fixing structural errors of predicted structures through direct coordinate refinement. Availability and implementation The source code of ATOMRefine is available in the GitHub repository (https://github.com/BioinfoMachineLearning/ATOMRefine). All the required data for training and testing are available at https://doi.org/10.5281/zenodo.6944368.
59 BASIC BIOLOGICAL SCIENCES↗