Search NASA⌕ Search

SEARCH · Search NASA

Results for “Protein modeling”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Understanding the selectivity of nonsteroidal anti-inflammatory drugs for cyclooxygenases using quantum crystallography and electrostatic interaction energy

Quantum crystallography methods have been employed to investigate complex formation between nonsteroidal anti-inflammatory drugs (NSAIDs) and cyclooxygenase (COX) enzymes, with particular focus on the COX-1 and COX-2 isoforms. This study analyzed the electrostatic interaction energies of selected NSAIDs (flurbiprofen, ibuprofen, meloxicam and celecoxib) with the active sites of COX-1 and COX-2, revealing significant differences in binding profiles. Flurbiprofen exhibited the strongest interactions with both COX-1 and COX-2, indicating its potent binding affinity. Celecoxib and meloxicam showed a preference for COX-2, consistent with their known selectivity for this isoform, while ibuprofen showed comparable interaction energies with both isoforms, reflecting its nonselective inhibition pattern. Key amino-acid residues, including Arg120, Arg/His513 and Tyr355, were identified as critical determinants of NSAID selectivity and binding affinity. The findings highlight the complex interplay between interaction energy and selectivity, suggesting that while electrostatic interactions play a fundamental role, additional factors such as enzyme dynamics and the hydrophobic effect also contribute to the therapeutic efficacy and safety profiles of NSAIDs. These insights provide valuable guidance for the rational design of NSAIDs with enhanced therapeutic benefits and minimized adverse effects.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2026

smol_model_benchmark

Machine learning models for protein–small molecule binding prediction using graph (GAT), image-based (ResNet), and transformer-based SMILES approaches (ChemBERTa and MIST). The code includes training pipelines and experiments on the dataset from the BELKA Kaggle competition.

Gibson, Kaetlyn [Los Alamos National Lab]↗
DOE OSTI2025

Nitrogen limitation causes a seismic shift in redox state and phosphorylation of proteins implicated in carbon flux and lipidome remodeling in Rhodotorula toruloides

Background: Oleaginous yeast are prodigious producers of oleochemicals, offering alternative and secure sources for applications in foodstuff, skincare, biofuels, and bioplastics. Nitrogen starvation is the primary strategy used to induce oil accumulation in oleaginous yeast as part of a global stress response. While research has demonstrated that post-translational modifications (PTMs), including phosphorylation and protein cysteine thiol oxidation (redox PTMs), are involved in signaling pathways that regulate stress responses in metazoa and algae, their role in oleaginous yeast remain understudied and unexplored. Results: Towards linking the yeast oleaginous phenotype to protein function, we integrated lipidomics, redox proteomics, and phosphoproteomics to investigate Rhodotorula toruloides under nitrogen-rich and starved conditions over time. Our lipidomics results unearthed interactions involving sphingolipids and cardiolipins with ER stress and mitophagy. Our redox and phosphoproteomics data highlighted the roles of the AMPK, TOR, and calcium signaling pathways in regulation of lipogenesis, autophagy, and oxidative stress response. As a first, we also demonstrated that lipogenic enzymes including fatty acid synthase are modified as a consequence of shifts in cellular redox states due to nutrient availability. Conclusions: We conclude that lipid accumulation is largely a consequence of carbon rerouting and autophagy governed by changes to PTMs, and not increases in the abundance of enzymes involved in central carbon metabolism and fatty acid biosynthesis. Our systems-level approach sets the stage for acquiring multidimensional data sets for protein structural modeling and predicting the functional relevance of PTMs using Artificial Intelligence/Machine Learning (AI/ML). Coupled to those bioinformatics approaches, the putative PTM switches that we delineate will enable advanced metabolic engineering strategies to decouple lipid accumulation from nitrogen limitation.

Lipid Signalling↗
DOE OSTI2026

Genomic-based biosurveillance for avian influenza: whole genome sequencing from wild mallards sampled during autumn migration in 2022–2023 reveals a high co-infection rate on migration stopover site in Georgia

The Caucasus region, including Georgia, is an important intersection for migratory waterbirds, offering potential for avian influenza virus (AIV) transmission between populations from different geographic areas. In 2022 and 2023, wild ducks were sampled during autumn migration events in Georgia to study the genetic relationships and molecular characteristics of influenza strains. Sequencing and phylogenetic analysis were used to compare the sampled strains to reference sequences from Africa, Asia, and Europe, allowing assessment of genetic relationships and virus transmission between migratory birds. Protein language modeling identified potential co-infections. Of 225 duck samples, 128 tested positive for the influenza M gene. 55 influenza-positive samples underwent whole-genome sequencing, revealing significant diversity. Analysis of the hemagglutinin (HA) segment showed notable differences among subtypes. Most samples were H6N1 and H6N6, but co-infections with combinations like H6H3, N8N1, N6H9, N2N6, and H9H6/N1N2 were also identified. These findings demonstrate the high variability of influenza viruses in migratory waterbirds in Georgia, including a notable rate of co-infections. Some samples exhibited uncommon genetic characteristics compared to other strains from the same year, suggesting Georgia’s role as a mixing vessel for influenza viruses. This facilitates reassortment during co-infections and contributes to the genetic diversity observed across flyways.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

CryoSegNet: accurate cryo-EM protein particle picking by integrating the foundational AI image segmentation model and attention-gated U-Net

Picking protein particles in cryo-electron microscopy (cryo-EM) micrographs is a crucial step in the cryo-EM-based structure determination. However, existing methods trained on a limited amount of cryo-EM data still cannot accurately pick protein particles from noisy cryo-EM images. The general foundational artificial intelligence–based image segmentation model such as Meta’s Segment Anything Model (SAM) cannot segment protein particles well because their training data do not include cryo-EM images. Here, we present a novel approach (CryoSegNet) of integrating an attention-gated U-shape network (U-Net) specially designed and trained for cryo-EM particle picking and the SAM. The U-Net is first trained on a large cryo-EM image dataset and then used to generate input from original cryo-EM images for SAM to make particle pickings. CryoSegNet shows both high precision and recall in segmenting protein particles from cryo-EM micrographs, irrespective of protein type, shape and size. On several independent datasets of various protein types, CryoSegNet outperforms two top machine learning particle pickers crYOLO and Topaz as well as SAM itself. The average resolution of density maps reconstructed from the particles picked by CryoSegNet is 3.33 Å, 7% better than 3.58 Å of Topaz and 14% better than 3.87 Å of crYOLO. It is publicly available at https://github.com/jianlin-cheng/CryoSegNet

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

PRIME: Protein Representation Inference for Mutation Evaluation

Protein language machine learning models built upon existing ESM-2 model developed by Evolutionary Scale (evolutionaryscale.ai) and an in-house protein language model based on the BERT model developed by Google. The code also includes model training scripts and saved checkpoints from our own training using publicly available SARS-CoV-2 protein sequences.

Gibson, Kaetlyn [Los Alamos National Lab]↗
DOE OSTI2025

Enzyme property prediction using artificial intelligence

Artificial intelligence (AI)-driven enzyme property prediction enables rapid discovery and engineering of enzymes for a wide range of biotechnological and therapeutic applications. Here, we first introduce the key components in AI model development, including enzyme datasets, protein representation methods, and model architectures. We then highlight a variety of AI tools developed for the prediction of enzyme properties and functional annotations, including enzyme structure, kinetic parameters, substrate specificity, thermostability, solubility, Enzyme Commission number, and Gene Ontology term. Moreover, we describe representative downstream applications enabled by these AI tools. Finally, we discuss some challenges and opportunities as well as future prospects.

Yuan, Le [University of Illinois at Urbana-Champai↗
DOE OSTI2025

Enzymatic carbon–fluorine bond cleavage by human gut microbes

Fluorinated compounds are used for agrochemical, pharmaceutical, and numerous industrial applications, resulting in global contamination. In many molecules, fluorine is incorporated to enhance the half-life and improve bioavailability. Fluorinated compounds enter the human body through food, water, and xenobiotics including pharmaceuticals, exposing gut microbes to these substances. The human gut microbiota is known for its xenobiotic biotransformation capabilities, but it was not previously known whether gut microbial enzymes could break carbon-fluorine bonds, potentially altering the toxicity of these compounds. Here, through the development of a rapid, miniaturized fluoride detection assay for whole-cell screening, we identified active gut microbial defluorinases. We biochemically characterized enzymes from diverse human gut microbial classes including Clostridia, Bacilli, and Coriobacteriia, with the capacity to hydrolyze (di)fluorinated organic acids and a fluorinated amino acid. Whole-protein alanine scanning, molecular dynamics simulations, and chimeric protein design enabled the identification of a disordered C-terminal protein segment involved in defluorination activity. Domain swapping exclusively of the C-terminus conferred defluorination activity to a nondefluorinating dehalogenase. To advance our understanding of the structural and sequence differences between defluorinating and nondefluorinating dehalogenases, we trained machine learning models which identified protein termini as important features. Models trained on 41-amino acid segments from protein C termini alone predicted defluorination activity with 83% accuracy (compared to 95% accuracy based on full-length protein features). This work is relevant for therapeutic interventions and environmental and human health by uncovering specificity-determining signatures of fluorine biochemistry from the gut microbiome.

Probst, Silke I↗
DOE OSTI2024

Specialization Restricts the Evolutionary Paths Available to Yeast Sugar Transporters

Functional innovation at the protein level is a key source of evolutionary novelties. The constraints on functional innovations are likely to be highly specific in different proteins, which are shaped by their unique histories and the extent of global epistasis that arises from their structures and biochemistries. These contextual nuances in the sequence–function relationship have implications both for a basic understanding of the evolutionary process and for engineering proteins with desirable properties. Here, we have investigated the molecular basis of novel function in a model member of an ancient, conserved, and biotechnologically relevant protein family. These Major Facilitator Superfamily sugar porters are a functionally diverse group of proteins that are thought to be highly plastic and evolvable. By dissecting a recent evolutionary innovation in an α-glucoside transporter from the yeast Saccharomyces eubayanus, we show that the ability to transport a novel substrate requires high-order interactions between many protein regions and numerous specific residues proximal to the transport channel. To reconcile the functional diversity of this family with the constrained evolution of this model protein, we generated new, state-of-the-art genome annotations for 332 Saccharomycotina yeast species spanning ~400 My of evolution. By integrating phylogenetic and phenotypic analyses across these species, we show that the model yeast α-glucoside transporters likely evolved from a multifunctional ancestor and became subfunctionalized. The accumulation of additive and epistatic substitutions likely entrenched this subfunction, which made the simultaneous acquisition of multiple interacting substitutions the only reasonably accessible path to novelty.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

SLAB: simultaneous labeling and binding affinity prediction for protein–ligand structures

Machine learning models are often used as scoring functions to predict the binding affinity of a protein–ligand complex. These models are trained with limited amounts of data with experimentally measured binding affinity values. A large number of compounds are labeled inactive through single-concentration screens without measuring binding affinities. These inactive compounds, along with the active ones, can be used to train binary classification models, while regression models are trained using compounds with binding affinities only. However, the classification and regression tasks are often handled separately, without sharing the learned feature representations. In this paper, we propose a novel model architecture that jointly performs regression and classification objectives, aiming to maximize data utilization and improve predictive performance by leveraging two complementary tasks. In our setup, the regression yields the binding affinity, whereas the classification task yields the label as active or inactive. We demonstrate our method using PDBbind, the standard 3D structure database, as well as a dataset of flavivirus protease compounds with binding affinity data. Our experiments show that the new joint training strategy improves the accuracy of the model, increasing applicability in various practical drug screening scenarios.

Biological and medical sciences↗
DOE OSTI2025

Energetic and structural control of polyspecificity in a multidrug transporter

Multidrug efflux pumps are dynamic molecular machines that drive antibiotic resistance by harnessing ion gradients to export chemically diverse substrates. Despite their clinical importance, the molecular principles underlying multidrug promiscuity and energy efficiency remain poorly understood. Using multiparametric deep mutational scanning across eight substrates and two energy conditions, we deconvolute the contributions of substrate recognition, energetic coupling, and protein stability, providing an integrated, high-resolution view of multidrug transport. We find that substrate specificity arises from a distributed network of residues extending beyond the binding site, with mutations that reshape binding, coupling, conformational flexibility, and membrane interactions. Further, we apply a pH-based selection scheme to measure the effect of mutation on pH-dependent transport efficiency. By integrating these data, we reveal a fundamental relationship between efficiency and promiscuity: Highly efficient variants exhibit broad substrate profiles, while inefficient variants are narrower. In conclusion, these findings establish a direct link between energy coupling and polyspecificity, uncovering the biochemical logic underlying multidrug transport.

Biological Sciences↗
DOE OSTI2026

Correlating Protein Dynamics and Catalytic Activity of a Model Hydrogenase Using Paramagnetic and Biological Nuclear Magnetic Resonance Spectroscopy

Rational catalyst design remains a significant challenge, with electronic structure, steric, and electrostatic effects known to contribute to activity. Recently, dynamics has been recognized as another factor that impacts catalysis, though identifying and predicting these effects has remained out of reach. Nickel-substituted rubredoxin (NiRd), a protein-based mimic of a hydrogenase enzyme, serves as a model catalytic system in which dynamics can be systematically investigated with respect to activity. While over 30 secondary-sphere mutants of NiRd have been shown to be catalytically active, no significant correlation was observed between the rates and catalytic overpotential or electronic structure, prompting questions about the protein-derived factors that modulate activity. Here, in this work, NMR spectroscopy was used to investigate the roles of substrate accessibility, protein dynamics, and protein stability in controlling catalysis. Significant paramagnetic effects from the nickel center (S = 1) isolate the methylene proton resonances of the metal-coordinating cysteine residues. The sensitivity of resonance positions and linewidths to local environment offers an opportunity to study dynamical molecular changes around the metal center with high resolution. Machine learning algorithms were employed to identify correlations between the catalytic activity and the paramagnetic NMR spectra. These analyses revealed spectroscopic features of specific cysteine protons that report on catalytic overpotential and increased turnover rates, which are further supported by the results obtained using high-field NMR techniques. Collectively, these studies indicate the potential for multifrequency NMR techniques to resolve key contributors to catalytic activity and highlight the importance of local and outer-sphere dynamics.

Protein Engineering↗
DOE OSTI2025

Predicting metal-binding proteins and structures through integration of evolutionary-scale and physics-based modeling

Metals are essential elements in all living organisms, binding to approximately 50% of proteins. They serve to stabilize proteins, catalyze reactions, regulate activities, and fulfill various physiological and pathological functions. While there have been many advancements in determining the structures of protein-metal complexes, numerous metal-binding proteins still need to be identified through computational methods and validated through experiments. Here, to address this need, we have developed the ESMBind workflow, which combines evolutionary scale modeling (ESM) for metal-binding prediction and physics-based protein-metal modeling. Our approach utilizes the ESM-2 and ESM-IF models to predict metal-binding probability at the residue level. In addition, we have designed a metal-placement method and energy minimization technique to generate detailed 3D structures of protein-metal complexes. Our workflow outperforms other models in terms of residue and 3D-level predictions. To demonstrate its effectiveness, we applied the workflow to 142 uncharacterized fungal pathogen proteins and predicted metal-binding proteins involved in fungal infection and virulence.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Designing Peptide Fossils That Model the Evolution of the Bacterial Ferredoxin Fold

Electron transfer coupled to redox chemistry is at the heart of metabolism. The proteins responsible for moving electrons (protein electron carriers) must have emerged at the origin of life. The small iron–sulfur-binding bacterial ferredoxins were likely among these first proteins. Embedded within the ferredoxin sequence and structure is a symmetry that points to an ancient gene duplication event. Little is understood about the nature of ferredoxins prior to this duplication event or what environmental factors may have driven the selection for more complex forms. The deep-time molecular history of ferredoxins goes back billions of years and cannot be reconstructed by phylogenetic analyses based on amino acid sequences. Here, we use structure-guided protein design to model a fossil half-ferredoxin stage in the evolution of this fold, the semidoxins, and their symmetric full-length counterparts, the symdoxins. Semidoxin designs homodimerize, exhibiting structural, thermodynamic, and electrochemical behaviors in most cases identical to cognate symdoxins. However, the semi- and symdoxin fossil stages behave differently when incorporated into an in vivo electron transfer complementation assay. Both can support bacterial growth dependent on protein expression. Growth rates of bacteria expressing the semidoxins are much more sensitive to oxygen than those of bacteria expressing symdoxins. Motivated by the in vivo functionality of designed semidoxins, we identified putative naturally occurring semidoxins in extant anaerobic microorganisms. This is consistent with the observed in vivo oxygen sensitivity of the semidoxin designs. One natural semidoxin is shown to be folded and redox active. However, it exists as a mixture of monomers and dimers, suggesting a potential connection between semidoxins and even simpler single iron–sulfur cluster-binding peptides.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Q -score as a reliability measure for protein, nucleic acid and small-molecule atomic coordinate models derived from 3DEM maps

Atomic coordinate models are important for the interpretation of 3D maps produced with cryoEM and cryoET (3D electron microscopy; 3DEM). In addition to visual inspection of such maps and models, quantitative metrics can inform about the reliability of the atomic coordinates, in particular how well the model is supported by the experimentally determined 3DEM map. A recently introduced metric, Q-score, was shown to correlate well with the reported resolution of the map for well fitted models. Here, we present new statistical analyses of Q-score based on its application to ∼10 000 maps and models archived in the EMDB (Electron Microscopy Data Bank) and PDB (Protein Data Bank). Further, we introduce two new metrics based on Q-score to represent each map and model relative to all entries in the EMDB and those with similar resolution. We explore through illustrative examples of proteins, nucleic acids and small molecules how Q-scores can indicate whether the atomic coordinates are well fitted to 3DEM maps and also whether some parts of a map may be poorly resolved due to factors such as molecular flexibility, radiation damage and/or conformational heterogeneity. These examples and statistical analyses provide a basis for how Q-scores can be interpreted effectively in order to evaluate 3DEM maps and atomic coordinate models prior to publication and archiving.

B factors↗
DOE OSTI2024

Automated AI-driven Molecular Design for Therapeutic Discovery

In recent years, artificial intelligence and machine learning (AI/ML) approaches have revolutionized the process of designing new therapeutics, enabling scientists to rapidly respond to emerging threats from various pathogens. A prime example is the SARS-CoV-2 main protease, a key target for the development of antiviral inhibitors. In this study, we employed a novel, integrated approach that combines AI-driven iterative design of inhibitor candidates, screening based on physio-chemical properties and toxicity, physics-based computational modeling of protein-inhibitor interactions, and AI-assisted analysis of Native MS biophysical assay and characterization of designed candidates. Our deep learning 3D-scaffold model, which uses an input scaffold as a starting point, generated tens of thousands of compounds while preserving the key scaffold. To optimize these candidates, we calculated a comprehensive set of 136 descriptors, including both 2D and 3D molecular features, for compounds targeting the SARS-CoV-2 Main protease (Mpro) and a neurodegenerative disease-associated protein, cyclophilin (Cyp). The generated compounds were initially filtered based on their properties and then ranked according to their predicted binding affinity using our automated modeling and ML methods. Experimental validation of the Mpro candidates showing inhibitory activity demonstrates that our workflow can expedite the therapeutic discovery.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Simultaneous enhancement of multiple functional properties using evolution-informed protein design

Abstract A major challenge in protein design is to augment existing functional proteins with multiple property enhancements. Altering several properties likely necessitates numerous primary sequence changes, and novel methods are needed to accurately predict combinations of mutations that maintain or enhance function. Models of sequence co-variation (e.g., EVcouplings), which leverage extensive information about various protein properties and activities from homologous protein sequences, have proven effective for many applications including structure determination and mutation effect prediction. We apply EVcouplings to computationally design variants of the model protein TEM-1 β -lactamase. Nearly all the 14 experimentally characterized designs were functional, including one with 84 mutations from the nearest natural homolog. The designs also had large increases in thermostability, increased activity on multiple substrates, and nearly identical structure to the wild type enzyme. This study highlights the efficacy of evolutionary models in guiding large sequence alterations to generate functional diversity for protein design applications.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

An essential and highly selective protein import pathway encoded by nucleus-forming phage

Targeting proteins to specific subcellular destinations is essential in prokaryotes, eukaryotes, and the viruses that infect them. Chimalliviridae phages encapsulate their genomes in a nucleus-like replication compartment composed of the protein chimallin (ChmA) that excludes ribosomes and decouples transcription from translation. These phages selectively partition proteins between the phage nucleus and the bacterial cytoplasm. Currently, the genes and signals that govern selective protein import into the phage nucleus are unknown. Here, we identify two components of this protein import pathway: a species-specific surface-exposed region of a phage intranuclear protein required for nuclear entry and a conserved protein, PicA (Protein importer of chimalliviruses A), that facilitates cargo protein trafficking across the phage nuclear shell. We also identify a defective cargo protein that is targeted to PicA on the nuclear periphery but fails to enter the nucleus, providing insight into the mechanism of nuclear protein trafficking. Using CRISPRi-ART protein expression knockdown of PicA, we show that PicA is essential early in the chimallivirus replication cycle. Together, our results allow us to propose a multistep model for the Protein Import Chimallivirus pathway, where proteins are targeted to PicA by amino acids on their surface and then licensed by PicA for nuclear entry. The divergence in the selectivity of this pathway between closely related chimalliviruses implicates its role as a key player in the evolutionary arms race between competing phages and their hosts.

59 BASIC BIOLOGICAL SCIENCES↗