Search NASA⌕ Search

SEARCH · Search NASA

Results for “Activation function”

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 433 records · Page 24
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

SANE: strategic autonomous non-smooth exploration for multiple optima discovery in multi-modal and non-differentiable black-box functions

Both computational and experimental material discovery bring forth the challenge of exploring multidimensional and multimodal parameter spaces, such as phase diagrams of Hamiltonians with multiple interactions, composition spaces of combinatorial libraries, material structure image spaces, and molecular embedding spaces. Often these systems are black-boxes and time-consuming to evaluate, which resulted in strong interest towards active learning methods such as Bayesian optimization (BO). However, these systems are often noisy which make the black box function severely multi-modal and non-differentiable, where a vanilla BO can get overly focused near a single or faux optimum, deviating from the broader goal of scientific discovery. To address these limitations, here we developed Strategic Autonomous Non-Smooth Exploration (SANE) to facilitate an intelligent Bayesian optimized navigation with a proposed cost-driven probabilistic acquisition function to find multiple global and local optimal regions, avoiding the tendency to becoming trapped in a single optimum. To distinguish between a true and false optimal region due to noisy experimental measurements, a human (domain) knowledge driven dynamic surrogate gate is integrated with SANE. We implemented the gate-SANE into pre-acquired piezoresponse spectroscopy data of a ferroelectric combinatorial library with high noise levels in specific regions, and piezoresponse force microscopy (PFM) hyperspectral data. SANE demonstrated better performance than classical BO to facilitate the exploration of multiple optimal regions and thereby prioritized learning with higher coverage of scientific values in autonomous experiments. Our work showcases the potential application of this method to real-world experiments, where such combined strategic and human intervening approaches can be critical to unlocking new discoveries in autonomous research.

Biswas, Arpan [University of Tennessee, Knoxville,↗
DOE OSTI2025

Prediction of α $IIb$ $β$ 3 integrin structures along its minimum free energy activation pathway

The adhesion protein integrin is a transmembrane heterodimer that plays a pivotal role in cellular processes such as cell signaling and cell migration. To execute its function, integrin undergoes extensive conformational changes from a bent-closed to an extended-open state. Resolving the structures across these changes remains a challenge with both experimental and computational methods, but it is crucial for understanding the activation mechanism of integrin. We address this challenge for the platelet integrin α IIb β 3 by employing finite temperature string method with structures of the images along the initial guess path generated by a multiscale data-driven framework. The full-length all-atom structures along the resulting minimum free energy path between the inactive bent-closed and active extended-open states of α IIb β 3 integrin are consistent with a variety of experimentally resolved structures. Changes in these predicted structures along the path show that the extension and separation of the α and β subunits from the bent-closed to the extended-open state require correlated movements between the subdomain pairs in α IIb β 3 . Furthermore, these results provide new insights into integrin activation mechanism, and the predicted structures have potential applications in guiding the design of integrin-targeting therapeutics.

Dasetty, Siva [University of Chicago, IL (United S↗
DOE OSTI2024

Structural dynamics at cytosolic interprotomer interfaces control gating of a mammalian TRPM5 channel

The transient receptor potential melastatin (TRPM) tetrameric cation channels are involved in a wide range of biological functions, from temperature sensing and taste transduction to regulation of cardiac function, inflammatory pain, and insulin secretion. The structurally conserved TRPM cytoplasmic domains make up >70 % of the total protein. To investigate the mechanism by which the TRPM cytoplasmic domains contribute to gating, we employed electrophysiology and cryo-EM to study TRPM5—a channel that primarily relies on activation via intracellular Ca 2+ . Here, we show that activation of mammalian TRPM5 channels is strongly altered by Ca 2+ -dependent desensitization. Structures of rat TRPM5 identify a series of conformational transitions triggered by Ca 2+ binding, whereby formation and dissolution of cytoplasmic interprotomer interfaces appear to control activation and desensitization of the channel. This study shows the importance of the cytoplasmic assembly in TRPM5 channel function and sets the stage for future investigations of other members of the TRPM family.

Science & Technology - Other Topics↗
DOE OSTI2024

Active learning for SNAP interatomic potentials via Bayesian predictive uncertainty

Bayesian inference with a simple Gaussian error model is used to efficiently compute prediction variances for energies, forces, and stresses in the linear SNAP interatomic potential. Here, the prediction variance is shown to have a strong correlation with the absolute error over approximately 24 orders of magnitude. Using this prediction variance, an active learning algorithm is constructed to iteratively train a potential by selecting the structures with the most uncertain properties from a pool of candidate structures. The relative importance of the energy, force, and stress errors in the objective function is shown to have a strong impact upon the trajectory of their respective net error metrics when running the active learning algorithm. Batched training of different batch sizes is also tested against singular structure updates, and it is found that batches can be used to significantly reduce the number of retraining steps required with only minor impact on the active learning trajectory.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Pyrolysis in Molten Salts Converts Plastics into a Mesoporous Electrocatalyst with a High Density of Atomic Fe–N–C Active Site

Upcycling plastic waste into value-added products is a promising strategy for both economic and environmental sustainability. However, achieving control over structure–property correlations during the upcycling process remains challenging. Here, in this study, we investigate the modulation of active site formation during pyrolytic conversion of plastics to carbon by composition-controlled molten salts. Using melamine formaldehyde (MF) foams as a precursor, we demonstrate that their pyrolysis in eutectic chloride molten salts with Fe 2+ , Na + , and K + ions directs the carbonization process toward the formation of mesoporous functional carbon enriched with atomically dispersed Fe–N–C sites, which are well-known for their catalytic activity in oxygen reduction reaction (ORR). Compared to pyrolysis in the FeCl 2 -only salt, the incorporation of alkali metal ions (Na + , K + ) in the eutectic mixture facilitates nitrogen retention in the carbon matrix and modulates iron speciation. This synergistic environment promotes a higher density and more uniform dispersion of Fe–N–C moieties within the carbon matrix. Consequently, the resulting catalyst exhibits a high surface area, hierarchical porosity, and improved electrochemical properties. When tested as an ORR catalyst, the Fe–N–C-enriched catalyst delivers a high half-wave potential (E 1/2 = 0.875 V vs RHE) with nearly exclusive 4e – pathway. These properties are comparable to those of commercial Pt/C catalysts, along with improved stability.

Fe−N−C active sites↗
DOE OSTI2025

An elastin-like polymer targeting vascular endothelial growth factor receptor-1 reduces survival in serum-starved endothelial cells

Peptides often exhibit biological activity that depends on the context in which they are displayed and delivered. Understanding and controlling these contextual effects on peptide function is critical for designing targeted and responsive peptide-based biomaterials and therapeutics. Genetically engineered protein polymers such as elastin-like polypeptides (ELPs) can incorporate bioactive peptide motifs and are attractive candidates for biomaterials used in tissue engineering and targeted drug delivery. They also present an opportunity for investigating and modulating cell signaling pathways by presenting a peptide ligand in various defined chemical and physical environments. Vascular endothelial growth factor receptor-1 (VEGFR1) signaling plays important and complex roles in cell survival and angiogenesis, but polymeric materials that interact with this signaling axis are scarce. In this study, a novel genetically engineered elastin-like polymer that targets VEGFR1 is characterized. This polymer, termed R1B-ELP, binds to human endothelial cells in a manner dependent on its VEGFR1-targeting motif and, based on cell proliferation and cytotoxicity assays, demonstrates activity consistent with disrupting pro-survival signaling necessary for endothelial cell function under conditions of environmental stress. Notably, these findings indicate that ELP fusion alters the functional behavior of the targeting peptide. Modulators of VEGFR1 signaling have potential applications in basic studies of angiogenesis as well as in therapeutic applications targeting vascular or inflammatory diseases.

36 MATERIALS SCIENCE↗
DOE OSTI2025

The molecular basis of Human FN3K mediated phosphorylation of glycated substrates

Abstract Glycation, a non-enzymatic post-translational modification occurring on proteins, can be actively reversed via site-specific phosphorylation of the fructose-lysine moiety by FN3K kinase, to impact the cellular function of the target protein. A regulatory axis between FN3K and glycated protein targets has been associated with conditions like diabetes and cancer. However, the molecular basis of this relationship has not been explored so far. Here, we determined a series of crystal structures of HsFN3K in the apo-state, and in complex with different nucleotide analogs together with a sugar substrate mimic to reveal the features important for its kinase activity and substrate recognition. Additionally, the dynamics in sugar substrate binding during the kinase catalytic cycle provide important mechanistic insights into HsFN3K function. Our structural work provides the molecular basis for rational small molecule design targeting FN3K.

Science & Technology - Other Topics↗
DOE OSTI2026

Characterizing Defect Dynamics in Silicon Carbide Using Symmetry-Adapted Collective Variables and Machine Learning Interatomic Potentials

Silicon carbide (SiC) divacancies are attractive candidates for spin-defect qubits possessing long coherence times and optical addressability. The high activation barriers associated with SiC defect formation and motion pose challenges for their study by first-principles molecular dynamics. In this work, we develop and deploy machine learning interatomic potentials (MLIPs) to accelerate defect dynamics simulations while retaining ab initio accuracy. We employ an active learning strategy comprising symmetry-adapted collective variable discovery and enhanced sampling to compile configurationally diverse training data, calculation of energies and forces using density functional theory (DFT), and training of an E(3)-equivariant MLIP based on the Allegro model. Here, the trained MLIP reproduces DFT-level accuracy in defect transition activation free energy barriers, enables the efficient and stable simulation of multidefect 216-atom supercells, and permits an analysis of the temperature dependence of defect thermodynamic stability and formation/annihilation kinetics to propose an optimal annealing temperature to maximally stabilize VV divacancies.

Computer simulations↗
DOE OSTI2024

Fragment-based initialization for quantum subspace methods

Here, we present a novel quantum-classical algorithm called LAS-QKSD for multireference systems, by combining a classical localized active space (LAS) fragment-based multireference algorithm with the quantum Krylov subspace diagonalization (QKSD) method for quantum computers. The algorithm uses wave function information from a LAS self-consistent field (LASSCF) calculation to prepare an initial state with better overlap with the target ground state than the Hartree-Fock state. This is coupled with the use of QKSD to ultimately converge to the exact energy, providing faster convergence than starting from the Hartree-Fock state. Fragmentation has the two-fold benefit of fewer configurations on the classical side of the algorithm as well as fewer state preparation gates on the quantum side. First, we compare the LAS-QKSD method to the classical LASSCF method and to QKSD with a Hartree-Fock initial state. We then examine ways to load the LASSCF wave function using direct initialization and a QKSD-motivated spectral filtering approach. Finally, using a bimetallic complex, we show that the LAS-QKSD method is an efficient alternative to highly expensive complete active space SCF (CASSCF) calculations on strongly correlated systems.

D'Cunha, Ruhee↗
DOE OSTI2026

Activation dynamics of a water-soluble human mu-opioid receptor

The mu-opioid receptor (MOR), a class A G protein-coupled receptor mediates opioid analgesia and remains a central target for pain therapeutics. While crystal structures of MOR exist, they provide limited insight into the receptor’s dynamic conformational landscape underlying function. Here, we engineered a thermostable water-soluble MOR variant (wsMOR) that retains native-like ligand-binding and activation dynamics. This variant enables high-yield production and detailed solution-phase structural studies that are challenging with membrane-embedded MOR, providing a valuable tool for studying receptor activation and aqueous-phase drug screening. Using a combined computational and experimental approach, we performed long-timescale all-atom molecular dynamics simulations together with neutron scattering and single-molecule FRET, revealing a structurally stable receptor with a diverse ensemble of conformations at different temporal resolutions. In the ligand-free state, wsMOR displayed high conformational flexibility, which decreased upon agonist binding, particularly in transmembrane helix 6, a hallmark of G protein-coupled receptor activation. Positive allosteric modulation and G protein binding further stabilized active-like states. These findings highlight wsMOR’s conformational plasticity across picosecond to millisecond timescales and provide a foundation for structure-guided development of next-generation opioid ligands with improved efficacy and safety.

E, Agyemang [University of Tennessee Knoxville]↗
DOE OSTI2024

Enhancing Nitrogen Activation in Electrochemical Reduction: The Role of Rare Earth Oxide Surface Configurations

The pursuit of sustainable ammonia synthesis has prompted the exploration of ambient electrochemical nitrogen reduction reaction (e-NRR) as an alternative to the energy-intensive Haber-Bosch process. Here, this study conducted a theoretical investigation into the use of rare earth oxide materials, specifically dysprosium oxide (Dy 2 O 3 ), as potential electrocatalysts for NRR. Utilizing spin-polarized density functional theory calculations, we explored the interaction between Dy 2 O 3 surfaces and nitrogen (N 2 ) molecules, examining the capability of Dy 2 O 3 to adsorb and activate N 2 under ambient conditions. The results indicate that Dy 2 O 3 surfaces exhibit diverse configurations and bonding environments, providing a variety of reactive sites that display different behaviors in N 2 adsorption and activation. The distinctive electronic structure and surface chemistry of a particular Dy 2 O 3 surface configuration were found to significantly enhance the activation of N 2 by promoting charge transfer, which facilitates the NRR process. This research provides deep insights into the mechanistic pathways of N 2 reduction over Dy 2 O 3 , highlighting the surface properties as pivotal in catalysis. These theoretical insights serve as a foundation for the development of novel rare earth-based electrocatalytic materials for efficient ambient e-NRR, potentially transforming ammonia production into a greener and more energy-efficient process.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

Machine learning guided selection of broad-spectrum epitope-specific functional antibodies for "Disease X"

Our project established and demonstrated a transfer learning framework that enables prediction of antibody–antigen interactions across related viruses. The approach focused on three major activities: 1. Conserved region and epitope identification – We compared viral protein structures and sequences to identify shared receptor-binding domains and neutralizing epitope regions across variants and related viruses. These conserved features formed the foundation for discovering broadly functional antibodies. 2. Machine learning model development – We built neural network–based models that integrate epitope features with antibody sequence information. Instead of relying solely on structural or physical properties, the models learned transferable patterns that describe antibody binding potential across different viral families. 3. Transfer learning and validation – Using SARS-CoV-2 and Ebola as source systems, we successfully transferred learned epitope features to predict antibody interactions for SARS CoV-1 and Marburg virus. Iterative cycles of dataset generation, retraining, and evaluation improved generalization and predictive power, ensuring the framework can adapt to new threats.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

The Role of Lignin Molecular Weight on Activated Carbon Pore Structure

Over the past decade, the production of biofuels from lignocellulosic biomass has steadily increased to offset the use of fuels from petroleum. To make biofuels cost-competitive, however, it is necessary to add value to the “ligno-” components (up to 30% by mass) of the biomass. The properties of lignin, in terms of molecular weight (MW), chemical functionality, and mineral impurities often vary from biomass source and biorefinery process, resulting in a challenging precursor for product development. Activated carbon (AC) is a feasible target for the lignin-rich byproduct streams because it can be made from nearly any biomass, and it has a market capacity large enough to use much of the lignin generated from the biorefineries. However, it is not known how the variability in the lignin affects the key properties of AC, because, until now, they could not be well controlled. In this work, various fractions of ultraclean (<0.6% ash) lignin are created with refined MW distributions using Aqueous Lignin Purification using Hot Agents (ALPHA) and used as precursors for AC. AC is synthesized via zinc chloride activation and characterized for pore structure and adsorption capacity. We show that AC surface area and the adsorption capacity increase when using lignin with increasing MW, and, furthermore, that reducing the mineral content of lignin can significantly enhance the AC properties. The surface area of the AC from the highest MW lignin can reach ~1830 m2/g (absorption capacity). Furthermore, single step activation carbonization using zinc chloride allows for minimal carbon burn off (<30%), capturing most of the lignin carbon compared to traditional burn off methods in biorefineries for heat generation.

Biochemistry & Molecular Biology↗
DOE OSTI2024

Insights into Prismatic Loop Formation in Irradiated Fe–Cr Alloys from Hypothesis-Driven Active Learning and Causal Analysis

Neutron and electron irradiation experimental studies conducted on body-centered cubic Fe and Fe–Cr alloys have established two prismatic dislocation loop populations, which have Burgers vectors of either a/2$\langle$111$\rangle$ or a$\langle$100$\rangle$. Here, the loop formation depends on factors such as dose (D), dose rate (D rt ), temperature (T), chromium content (Cr%), and other alloying elements. Hence, it is important to understand how irradiation-induced dislocation loops evolve conditional upon the loop characteristics, such as loop density (DD), average loop size d̅, and irradiation parameters (D, D rt , T, and irradiation type), which is still an active area of research. To understand these complex structure–property relationships, machine learning (ML) is employed in a three-step approach. This includes imputing missing data with a k-nearest neighbor, generating functionalized features, and assessing feature importance with random forest classification and regression. Physics-based features are incorporated in a hypothesis-driven active learning scheme to overcome data unavailability challenges. Insights obtained from ML models (i) to categorize dislocation loop types, show the highest correlation with d̅; (ii) Log(DD), obtained through mathematical formulations involving D, Cr%, d̅, and T (e.g., Log(DD) ~ D + exp(-Cr%) + 1/d̅ and log(DD) ~ D + exp(-Cr%) + 1/T). Hypothesis-driven active learning is able to predict Log(DD) in which the experimental date is not known. Causal models verify cause–effect relationships for dislocation loop classification and irradiation factors in FeCr alloys.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

A minimal complex of KHNYN and zinc-finger antiviral protein binds and degrades single-stranded RNA

Detecting viral infection is a key role of the innate immune system. The genomes of some RNA viruses have a high CpG dinucleotide content relative to most vertebrate cell RNAs, making CpGs a molecular marker of infection. The human zinc-finger antiviral protein (ZAP) recognizes CpG, mediates clearance of the foreign CpG-rich RNA, and causes attenuation of CpG-rich RNA viruses. While ZAP binds RNA, it lacks enzymatic activity that might be responsible for RNA degradation and thus requires interacting cofactors for its function. One of these cofactors, KHNYN, has a predicted nuclease domain. Using biochemical approaches, we found that the KHNYN NYN domain is a single-stranded RNA ribonuclease that does not have sequence specificity and digests RNA with or without CpG dinucleotides equivalently in vitro. We show that unlike most KH domains, the KHNYN KH domain does not bind RNA. Indeed, a crystal structure of the KH region revealed a double-KH domain with a negatively charged surface that accounts for the lack of RNA binding. Rather, the KHNYN C-terminal domain (CTD) interacts with the ZAP RNA-binding domain (RBD) to provide target RNA specificity. We define a minimal complex composed of the ZAP RBD and the KHNYN NYN-CTD and use a fluorescence polarization assay to propose a model for how this complex interacts with a CpG dinucleotide-containing RNA. In the context of the cell, this module would represent the minimum ZAP and KHNYN domains required for CpG-recognition and ribonuclease activity essential for attenuation of viruses with clusters of CpG dinucleotides.

Yeoh, Zoe C. (ORCID:0000000226949068)↗
DOE OSTI2025

Computing Reaction Kinetics with MC-PDFT–OPESf: Combining Multireference Electronic Structure Theory and Enhanced Sampling

Accurate rate constants are crucial for understanding and optimizing catalytic reactions mediated by enzymes, metalloproteins, and heterogeneous catalysts. These systems frequently present a dual computational challenge. Multiconfigurational reaction sites require multireference techniques for the accurate treatment of the electronic structure, and high activation barriers prevent efficient sampling of unbiased reactive transitions. In this work, we combine multiconfiguration pair-density functional theory (MC-PDFT) as an accurate and efficient multireference electronic structure method with on-the-fly probability-enhanced sampling flooding (OPESf) as an enhanced sampling method capable of accelerating reactive transitions. We demonstrate the approach on the Diels–Alder [4+2] cycloaddition between cis-butadiene and ethene as a reaction characterized by a large activation barrier and multireference character. MC-PDFT–OPESf provides reaction rates in agreement with experiments at a fraction of the computational cost required by conventional unbiased ab initio calculations. Here, we propose MC-PDFT–OPESf as an efficient approach for computing kinetics in strongly correlated molecular systems.

Chemical calculations↗
DOE OSTI2024

Interfacial Electrochemistry of Catalyst-Coordinated Graphene Nanoribbons

The immobilization of molecular electrocatalysts on conductive electrodes is an appealing strategy for enhancing their overall activity relative to those of analogous molecular compounds. Here, in this study, we report on the interfacial electrochemistry of self-assembled two-dimensional nanosheets of graphene nanoribbons (GNR-2DNS) and analogs containing a Rh-based hydrogen evolution reaction (HER) catalyst (RhGNR-2DNS) immobilized on conductive electrodes. Proton-coupled electron transfer (PCET) taking place at N-centers of the nanoribbons was utilized as an indirect reporter of the interfacial electric fields experienced by the monolayer nanosheet located within the electric double layer. The experimental Pourbaix diagrams were compared with a theoretical model, which derives the experimental Pourbaix slopes as a function of parameter f, a fraction of the interfacial potential drop experienced by the redox-active group. Interestingly, our study revealed that GNR-2DNS was strongly coupled to glassy carbon electrodes (f = 1), while RhGNR-2DNS was not (f = 0.15). We further investigated the HER mechanism by RhGNR-2DNS using electrochemical and X-ray absorption spectroelectrochemical methods and compared it to homogeneous molecular model compounds. RhGNR-2DNS was found to be an active HER electrocatalyst over a broader set of aqueous pH conditions than its molecular analogs. We find that the improved HER performance in the immobilized catalyst arises due to two factors. First, redox-active bipyrimidine-based ligands were shown to dramatically alter the activity of Rh sites by increasing the electron density at the active Rh center and providing RhGNR-2DNS with improved catalysis. Second, catalyst immobilization was found to prevent catalyst aggregation that was found to occur for the molecular analog in the basic pH. Overall, this study provides valuable insights into the mechanism by which catalyst immobilization can affect the overall electrocatalytic performance.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

The role of transition metal formyl intermediates in the reduction of CO and CO 2

Transition-metal formyl (metalloformyl) complexes occupy a central position in the activation of small molecules, particularly in the reduction of carbon monoxide (CO) and carbon dioxide (CO 2 ). This review examines five decades of progress in the synthesis of metalloformyl complexes and investigations into their structure and reactivity. The bonding in the M–CHO unit is best described as a resonance hybrid between a classical σ-bound formyl ligand and an oxycarbene-like electronic structure, which governs their distinctive spectroscopic signatures and versatile reactivity. Established synthetic routes are summarized, including pathways involving hydride addition and CO insertion, alongside a discussion of the thermodynamic and kinetic factors that control formyl stability. Decomposition pathways and Lewis-acid stabilization strategies are analyzed as key design principles for extending metalloformyl lifetimes under catalytic conditions. Particular attention is given to hydride transfer processes and the role of metalloformyl intermediates as both reactive substrates and hydride sources in reduction chemistry. Lastly, emerging catalytic strategies that exploit metalloformyl intermediates in CO 2 and CO reduction are evaluated, highlighting how control of hydricity, redox potential, and secondary-sphere interactions enables selective C—H bond formation under comparatively mild conditions. Collectively, these studies establish metalloformyl complexes as mechanistically informative and functionally relevant intermediates that bridge fundamental organometallic chemistry with modern approaches to small-molecule activation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗