Search NASA⌕ Search

SEARCH · Search NASA

Results for “Protein Structure Prediction”

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.

At least 199 records · Page 11

Functional protein mining with conformal guarantees

Molecular structure prediction and homology detection offer promising paths to discovering protein function and evolutionary relationships. However, current approaches lack statistical reliability assurances, limiting their practical utility for selecting proteins for further experimental and in-silico characterization. To address this challenge, we introduce a statistically principled approach to protein search leveraging principles from conformal prediction, offering a framework that ensures statistical guarantees with user-specified risk and provides calibrated probabilities (rather than raw ML scores) for any protein search model. Our method (1) lets users select many biologically-relevant loss metrics (i.e. false discovery rate) and assigns reliable functional probabilities for annotating genes of unknown function; (2) achieves state-of-the-art performance in enzyme classification without training new models; and (3) robustly and rapidly pre-filters proteins for computationally intensive structural alignment algorithms. Our framework enhances the reliability of protein homology detection and enables the discovery of uncharacterized proteins with likely desirable functional properties.

59 BASIC BIOLOGICAL SCIENCES↗

Structure of the C-terminal heme-binding domain of THAP domain containing protein 4 from Homo sapiens

The thanatos (the Greek god of death)-associated protein (THAP) domain is a sequence-specific DNA-binding domain that contains a C2-CH (Cys-Xaa{sub 2-4}-Cys-Xaa{sub 35-50}-Cys-Xaa{sub 2}-His) zinc finger that is similar to the DNA domain of the P element transposase from Drosophila. THAP-containing proteins have been observed in the proteome of humans, pigs, cows, chickens, zebrafish, Drosophila, C. elegans, and Xenopus. To date, there are no known THAP domain proteins in plants, yeast, or bacteria. There are 12 identified human THAP domain-containing proteins (THAP0-11). In all human THAP protein, the THAP domain is located at the N-terminus and is {approx}90 residues in length. Although all of the human THAP-containing proteins have a homologous N-terminus, there is extensive variation in both the predicted structure and length of the remaining protein. Even though the exact function of these THAP proteins is not well defined, there is evidence that they play a role in cell proliferation, apoptosis, cell cycle modulation, chromatin modification, and transcriptional regulation. THAP-containing proteins have also been implicated in a number of human disease states including heart disease, neurological defects, and several types of cancers. Human THAP4 is a 577-residue protein of unknown function that is proposed to bind DNA in a sequence-specific manner similar to THAP1 and has been found to be upregulated in response to heat shock. THAP4 is expressed in a relatively uniform manner in a broad range of tissues and appears to be upregulated in lymphoma cells and highly expressed in heart cells. The C-terminal domain of THAP4 (residues 415-577), designated here as cTHAP4, is evolutionarily conserved and is observed in all known THAP4 orthologs. Several single-domain proteins lacking a THAP domain are found in plants and bacteria and show significant levels of homology to cTHAP4. It appears that cTHAP4 belongs to a large class of proteins that have yet to be fully functionally characterized. On the basis of prior work, we predicted that cTHAP4 is composed of a heme-binding nitrobindin domain, making THAP4 the only human THAP protein predicted to bind a cofactor. Nitrobindin, a recently characterized protein from Arabidopsis thaliana, is structurally similar and exhibits nitric oxide (NO)-binding properties that resemble the heme-binding nitrophorins. Nitrophorins use a heme moiety to store, transport, and release NO in a pH-specific manner. Although the exact function of nitrobindin is not fully known, the similarities between the well-characterized nitrophorins imply a role in NO transport, sensing, or metabolism. To better elucidate the possible function of THAP4, we solved the hemebound structure of cTHAP4 to a resolution of 1.79 {angstrom}.

59 BASIC BIOLOGICAL SCIENCES↗

Deep learning-driven insights into super protein complexes for outer membrane protein biogenesis in bacteria

To reach their final destinations, outer membrane proteins (OMPs) of gram-negative bacteria undertake an eventful journey beginning in the cytosol. Multiple molecular machines, chaperones, proteases, and other enzymes facilitate the translocation and assembly of OMPs. These helpers usually associate, often transiently, forming large protein assemblies. They are not well understood due to experimental challenges in capturing and characterizing protein-protein interactions (PPIs), especially transient ones. Using AF2Complex, we introduce a high-throughput, deep learning pipeline to identify PPIs within the Escherichia coli cell envelope and apply it to several proteins from an OMP biogenesis pathway. Among the top confident hits obtained from screening ~1500 envelope proteins, we find not only expected interactions but also unexpected ones with profound implications. Subsequently, we predict atomic structures for these protein complexes. These structures, typically of high confidence, explain experimental observations and lead to mechanistic hypotheses for how a chaperone assists a nascent, precursor OMP emerging from a translocon, how another chaperone prevents it from aggregating and docks to a β-barrel assembly port, and how a protease performs quality control. This work presents a general strategy for investigating biological pathways by using structural insights gained from deep learning-based predictions.

60 APPLIED LIFE SCIENCES↗

A Computational Framework for Proteome-Wide Pursuit and Prediction of Metalloproteins using ICP-MS and MS/MS Data

Background: Metal-containing proteins comprise a diverse and sizable category within the proteomes of organisms, ranging from proteins that use metals to catalyze reactions to proteins in which metals play key structural roles. Unfortunately, reliably predicting that a protein will contain a specific metal from its amino acid sequence is not currently possible. We recently developed a generally-applicable experimental technique for finding metalloproteins on a genome-wide scale. Applying this metal-directed protein purification approach (ICPMS and MS/MS based) to the prototypical microbe Pyrococcus furiosus conclusively demonstrated the extent and diversity of the uncharacterized portion of microbial metalloproteomes since a majority of the observed metal peaks could not be assigned to known or predicted metalloproteins. However, even using this technique, it is not technically feasible to purify to homogeneity all metalloproteins in an organism. In order to address these limitations and complement the metal-directed protein purification, we developed a computational infrastructure and statistical methodology to aid in the pursuit and identification of novel metalloproteins. Results: We demonstrate that our methodology enables predictions of metal-protein interactions using an experimental data set derived from a chromatography fractionation experiment in which 870 proteins and 10 metals were measured over 2,589 fractions. For each of the 10 metals, cobalt, iron, manganese, molybdenum, nickel, lead, tungsten, uranium, vanadium, and zinc, clusters of proteins frequently occurring in metal peaks (of a specific metal) within the fractionation space were defined. This resulted in predictions that there are from 5 undiscovered vanadium- to 13 undiscovered cobalt-containing proteins in Pyrococcus furiosus. Molybdenum and nickel were chosen for additional assessment producing lists of genes predicted to encode metalloproteins or metalloprotein subunits, 22 for nickel including seven from known nickel-proteins, and 20 for molybdenum including two from known molybdo-proteins. The uncharacterized proteins are prime candidates for metal-based purification or recombinant approaches to validate these predictions. Conclusions: We conclude that the largely uncharacterized extent of native metalloproteomes can be revealed through analysis of the co-occurrence of metals and proteins across a fractionation space. This can significantly impact our understanding of metallobiochemistry, disease mechanisms, and metal toxicity, with implications for bioremediation, medicine and other fields.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

RosettaBackrub--a web server for flexible backbone protein structure modeling and design

The Rosetta-Backrub server (http://kortemmelab .ucsf.edu/backrub) implements the Backrub method, derived from observations of alternative conformations in high-resolution protein crystal structures, for flexible backbone protein modeling. Backrub modeling is applied to three related applications using the Rosetta program for structure prediction and design: (I) modeling of structures of point mutations, (II) generating protein conformational ensembles and designing sequences consistent with these conformations and (III) predicting tolerated sequences at protein–protein interfaces. The three protocols have been validated on experimental data. Starting from a user-provided single input protein structure in PDB format, the server generates near-native conformational ensembles. The predicted conformations and sequences can be used for different applications, such as to guide mutagenesis experiments, for ensemble-docking approaches or to generate sequence libraries for protein design.

59 BASIC BIOLOGICAL SCIENCES↗

Generative design of de novo proteins based on secondary-structure constraints using an attention-based diffusion model

We report two generative deep-learning models that predict amino acid sequences and 3D protein structures on the basis of secondary-structure design objectives via either the overall content or the per-residue structure. Both models are robust regarding imperfect inputs and offer de novo design capacity because they can discover new protein sequences not yet discovered from natural mechanisms or systems. The residue-level secondary-structure design model generally yields higher accuracy and more diverse sequences. These findings suggest unexplored opportunities for protein designs and functional outcomes within the vast amino acid sequences beyond known proteins. Our models, based on an attention-based diffusion model and trained on a dataset extracted from experimentally known 3D protein structures, offer numerous downstream applications in the conditional generative design of various biological or engineering systems. Future work could include additional conditioning and an exploration of other functional properties of the generated proteins for various properties beyond structural objectives.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Fingerprinting Interactions between Proteins and Ligands for Facilitating Machine Learning in Drug Discovery

Molecular recognition is fundamental in biology, underpinning intricate processes through specific protein–ligand interactions. This understanding is pivotal in drug discovery, yet traditional experimental methods face limitations in exploring the vast chemical space. Computational approaches, notably quantitative structure–activity/property relationship analysis, have gained prominence. Molecular fingerprints encode molecular structures and serve as property profiles, which are essential in drug discovery. While two-dimensional (2D) fingerprints are commonly used, three-dimensional (3D) structural interaction fingerprints offer enhanced structural features specific to target proteins. Machine learning models trained on interaction fingerprints enable precise binding prediction. Recent focus has shifted to structure-based predictive modeling, with machine-learning scoring functions excelling due to feature engineering guided by key interactions. Notably, 3D interaction fingerprints are gaining ground due to their robustness. Various structural interaction fingerprints have been developed and used in drug discovery, each with unique capabilities. This review recapitulates the developed structural interaction fingerprints and provides two case studies to illustrate the power of interaction fingerprint-driven machine learning. The first elucidates structure–activity relationships in β2 adrenoceptor ligands, demonstrating the ability to differentiate agonists and antagonists. The second employs a retrosynthesis-based pre-trained molecular representation to predict protein–ligand dissociation rates, offering insights into binding kinetics. Despite remarkable progress, challenges persist in interpreting complex machine learning models built on 3D fingerprints, emphasizing the need for strategies to make predictions interpretable. Binding site plasticity and induced fit effects pose additional complexities. Interaction fingerprints are promising but require continued research to harness their full potential.

3D structural interaction fingerprints↗

DBSI: DNA-binding site identifier

In this study, we present the DNA-Binding Site Identifier (DBSI), a new structure-based method for predicting protein interaction sites for DNA binding. DBSI was trained and validated on a data set of 263 proteins (TRAIN-263), tested on an independent set of protein-DNA complexes (TEST-206) and data sets of 29 unbound (APO-29) and 30 bound (HOLO-30) protein structures distinct from the training data. We computed 480 candidate features for identifying protein residues that bind DNA, including new features that capture the electrostatic microenvironment within shells near the protein surface. Our iterative feature selection process identified features important in other models, as well as features unique to the DBSI model, such as a banded electrostatic feature with spatial separation comparable with the canonical width of the DNA minor groove. Validations and comparisons with established methods using a range of performance metrics clearly demonstrate the predictive advantage of DBSI, and its comparable performance on unbound (APO-29) and bound (HOLO-30) conformations demonstrates robustness to binding-induced protein conformational changes. Finally, we offer our feature data table to others for integration into their own models or for testing improved feature selection and model training strategies based on DBSI.

59 BASIC BIOLOGICAL SCIENCES↗

Lysine carboxylation: unveiling a spontaneous post-translational modification

A computational method for the prediction of lysine carboxylation (KCX) in protein structures is described. The method accurately identifies misreported KCXs and predicts previously unknown KCX sites. The carboxylation of lysine residues is a post-translational modification (PTM) that plays a critical role in the catalytic mechanisms of several important enzymes. It occurs spontaneously under certain physicochemical conditions, but is difficult to detect experimentally. Its full impact is unknown. In this work, the signature microenvironment of lysine-carboxylation sites has been characterized. In addition, a computational method called Predictor of Lysine Carboxylation (PreLysCar) for the detection of lysine carboxylation in proteins with available three-dimensional structures has been developed. The likely prevalence of lysine carboxylation in the proteome was assessed through large-scale computations. The results suggest that about 1.3% of large proteins may contain a carboxylated lysine residue. This unexpected prevalence of lysine carboxylation implies an enrichment of reactions in which it may play functional roles. The results also suggest that by switching enzymes on and off under appropriate physicochemical conditions spontaneous PTMs may serve as an important and widely used efficient biological machinery for regulation.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

BIPSPI+: Mining Type-Specific Datasets of Protein Complexes to Improve Protein Binding Site Prediction

Computational approaches for predicting protein-protein interfaces are extremely useful for understanding and modelling the quaternary structure of protein assemblies. In particular, partner-specific binding site prediction methods allow delineating the specific residues that compose the interface of protein complexes. In recent years, new machine learning and other algorithmic approaches have been proposed to solve this problem. However, little effort has been made in finding better training datasets to improve the performance of these methods. With the aim of vindicating the importance of the training set compilation procedure, in this work we present BIPSPI+, a new version of our original server trained on carefully curated datasets that outperforms our original predictor. We show how prediction performance can be improved by selecting specific datasets that better describe particular types of protein interactions and interfaces (e.g. homo/hetero). In addition, our upgraded web server offers a new set of functionalities such as the sequence-structure prediction mode, hetero- or homo-complex specialization and the guided docking tool that allows to compute 3D quaternary structure poses using the predicted interfaces. BIPSPI+ is freely available at https://bipspi.cnb.csic.es.

59 BASIC BIOLOGICAL SCIENCES↗

Pairwise covariance adds little to secondary structure prediction but improves the prediction of non-canonical local structure

Amino acid sequence probability distributions, or profiles, have been used successfully to predict secondary structure and local structure in proteins. Profile models assume the statistical independence of each position in the sequence, but the energetics of protein folding is better captured in a scoring function that is based on pairwise interactions, like a force field. I-sites motifs are short sequence/structure motifs that populate the protein structure database due to energy-driven convergent evolution. Here we show that a pairwise covariant sequence model does not predict alpha helix or beta strand significantly better overall than a profile-based model, but it does improve the prediction of certain loop motifs. The finding is best explained by considering secondary structure profiles as multivariant, all-or-none models, which subsume covariant models. Pairwise covariance is nonetheless present and energetically rational. Examples of negative design are present, where the covariances disfavor non-native structures. Measured pairwise covariances are shown to be statistically robust in cross-validation tests, as long as the amino acid alphabet is reduced to nine classes. We present an updated I-sites local structure motif library and web server that provide sequence covariance information for all types of local structure in globular proteins.

59 BASIC BIOLOGICAL SCIENCES↗

Crystal Structure of the HEAT Domain from the Pre-mRNA Processing Factor Symplekin

The majority of eukaryotic pre-mRNAs are processed by 3'-end cleavage and polyadenylation, although in metazoa the replication-dependent histone mRNAs are processed by 3'-end cleavage but not polyadenylation. The macromolecular complex responsible for processing both canonical and histone pre-mRNAs contains the {approx} 1160-residue protein Symplekin. Secondary-structural prediction algorithms identified putative HEAT domains in the 300 N-terminal residues of all Symplekins of known sequence. The structure and dynamics of this domain were investigated to begin elucidating the role Symplekin plays in mRNA maturation. The crystal structure of the Drosophila melanogaster Symplekin HEAT domain was determined to 2.4 {angstrom} resolution with single-wavelength anomalous dispersion phasing methods. The structure exhibits five canonical HEAT repeats along with an extended 31-amino-acid loop (loop 8) between the fourth and fifth repeat that is conserved within closely related Symplekin sequences. Molecular dynamics simulations of this domain show that the presence of loop 8 dampens correlated and anticorrelated motion in the HEAT domain, therefore providing a neutral surface for potential protein-protein interactions. HEAT domains are often employed for such macromolecular contacts. The Symplekin HEAT region not only structurally aligns with several established scaffolding proteins, but also has been reported to contact proteins essential for regulating 3'-end processing. Together, these data support the conclusion that the Symplekin HEAT domain serves as a scaffold for protein-protein interactions essential to the mRNA maturation process.

36 MATERIALS SCIENCE↗

Omics-guided metabolic pathway discovery in plants: Resources, approaches, and opportunities

Plants produce a vast array of metabolites, the biosynthetic routes of which remain largely undetermined. Genome-scale enzyme and pathway annotations and omics technologies have revolutionized research to decrypt plant metabolism and produced a growing list of functionally characterized metabolic genes and pathways. However, what is known is still a tiny fraction of the metabolic capacity harbored by plants. Here, in this work, we review plant enzyme and pathway annotation resources and cutting-edge omics approaches to guide discovery and characterization of plant metabolic pathways. We also discuss strategies for improving enzyme function prediction by integrating protein 3D structure information and single cell omics. This review aims to serve as a primer for plant biologists to leverage omics datasets to facilitate understanding and engineering plant metabolism.

59 BASIC BIOLOGICAL SCIENCES↗

Machine learning sheds light on microbial dark proteins

In this article, metagenomics projects have revealed more than 8 billion non-redundant microbial protein sequences from across the Earth’s biosphere. Of these, 1.17 billion proteins do not have recognizable homologues in any of the more than 100,000 reference genomes available1. Understanding the function of these microbial proteins is a daunting task. Fortunately, machine learning has recently achieved unprecedented accuracy in modelling complex biological data and making predictions. At the forefront of these advancements are machine learning-based approaches that can confidently predict atomic-level protein structures for many (but not all) amino acid sequences.

59 BASIC BIOLOGICAL SCIENCES↗

Atypical Divergence of SARS-CoV-2 Orf8 from Orf7a within the Coronavirus Lineage Suggests Potential Stealthy Viral Strategies in Immune Evasion

Orf8, one of the most puzzling genes in the SARS lineage of coronaviruses, marks a unique and striking difference in genome organization between SARS-CoV-2 and SARS-CoV-1. Here, using sequence comparisons, we unequivocally reveal the distant sequence similarities between SARS-CoV-2 Orf8 with its SARS-CoV-1 counterparts and the X4-like genes of coronaviruses, including its highly divergent “paralog” gene Orf7a, whose product is a potential immune antagonist of known structure. Supervised sequence space walks unravel identity levels that drop below 10% and yet exhibit subtle conservation patterns in this novel superfamily, characterized by an immunoglobulin-like beta sandwich topology. We document the high accuracy of the sequence space walk process in detail and characterize the subgroups of the superfamily in sequence space by systematic annotation of gene and taxon groups. While SARS-CoV-1 Orf7a and Orf8 genes are most similar to bat virus sequences, their SARS-CoV-2 counterparts are closer to pangolin virus homologs, reflecting the fine structure of conservation patterns within the SARS-CoV-2 genomes. The divergence between Orf7a and Orf8 is exceptionally idiosyncratic, since Orf7a is more constrained, whereas Orf8 is subject to rampant change, a peculiar feature that may be related to hitherto-unknown viral infection strategies. Despite their common origin, the Orf7a and Orf8 protein families exhibit different modes of evolutionary trajectories within the coronavirus lineage, which might be partly attributable to their complex interactions with the mammalian host cell, reflected by a multitude of functional associations of Orf8 in SARS-CoV-2 compared to a very small number of interactions discovered for Orf7a.

60 APPLIED LIFE SCIENCES↗

Foldy v1

We created a cloud-based application for running AlphaFold2 and related structural programs called Foldy. Foldy is built on Helm / Kubernetes, which enables a facile production deployment and a scalable backend. Once set up by an institution, Foldy requires no software expertise from its users. It can be used to predict the structure of large proteins (up to 3000 amino acids), to visualize Pfam and antiSMASH annotations, and to perform ligand docking with Autodock Vina.

Roberts, Jacob↗

Improving the chances of successful protein structure determination with a random forest classifier

Using an extended set of protein features calculated separately for protein surface and interior, a new version of XtalPred based on a random forest classifier achieves a significant improvement in predicting the success of structure determination from the primary amino-acid sequence. Obtaining diffraction quality crystals remains one of the major bottlenecks in structural biology. The ability to predict the chances of crystallization from the amino-acid sequence of the protein can, at least partly, address this problem by allowing a crystallographer to select homologs that are more likely to succeed and/or to modify the sequence of the target to avoid features that are detrimental to successful crystallization. In 2007, the now widely used XtalPred algorithm [Slabinski et al. (2007 ▶), Protein Sci.16, 2472–2482] was developed. XtalPred classifies proteins into five ‘crystallization classes’ based on a simple statistical analysis of the physicochemical features of a protein. Here, towards the same goal, advanced machine-learning methods are applied and, in addition, the predictive potential of additional protein features such as predicted surface ruggedness, hydrophobicity, side-chain entropy of surface residues and amino-acid composition of the predicted protein surface are tested. The new XtalPred-RF (random forest) achieves significant improvement of the prediction of crystallization success over the original XtalPred. To illustrate this, XtalPred-RF was tested by revisiting target selection from 271 Pfam families targeted by the Joint Center for Structural Genomics (JCSG) in PSI-2, and it was estimated that the number of targets entered into the protein-production and crystallization pipeline could have been reduced by 30% without lowering the number of families for which the first structures were solved. The prediction improvement depends on the subset of targets used as a testing set and reaches 100% (i.e. twofold) for the top class of predicted targets.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗