Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Domain adaptation”

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 469 records · Page 26
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Mycobacterium tuberculosis CitA activity is modulated by cysteine oxidation and pyruvate binding

As an adaptation for survival during infection, Mycobacterium tuberculosis becomes dormant, reducing its metabolism and growth. Two types of citrate synthases have been identified in Mycobacterium tuberculosis, GltA2 and CitA. Previous work shows that overexpression of CitA, the secondary citrate synthase, stimulates the growth of Mycobacterium tuberculosis under hypoxic conditions without showing accumulation of triacylglycerols and makes mycobacteria more sensitive to antibiotics, suggesting that CitA may play a role as a metabolic switch during infection and may be an interesting TB drug target. To assess the druggability and possible mechanisms of targeting CitA with small-molecule compounds, the CitA crystal structure was solved to 2.1 Å by X-ray crystallography. The solved structure shows that CitA lacks an NADH binding site that would afford allosteric regulation, which is atypical of most citrate synthases. However, a pyruvate molecule is observed within the analogous domain, suggesting pyruvate may instead be the allosteric regulator for CitA. Further, the R149 and R153 residues forming the charged portion of the pyruvate binding pocket were mutated to glutamate and methionine, respectively, to assess the effect of mutations on activity. Protein thermal shift assay shows thermal stabilization of CitA in the presence of pyruvate compared to the two CitA variants designed to decrease pyruvate affinity. Solved crystal structures of both variants show no significant structural changes. However, the catalytic efficiency of the R153M variant increases by 2.6-fold. Additionally, we show that covalent modification of C143 of CitA by Ebselen completely arrests enzyme activity. Similar inhibition is observed using two spirocyclic Michael acceptor containing compounds, which inhibit CitA with IC app 50 values of 6.6 and 10.9 μM. A crystal structure of CitA modified by Ebselen was solved, but significant structural changes were lacking. Considering that covalent modification of C143 inactivates CitA and the proximity of C143 to the pyruvate binding site, this suggests that structural and/or chemical changes in this sub-domain are responsible for regulating CitA enzymatic activity.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Altered excitation energy transfer between phycobilisome and photosystems in the absence of ApcG, a small linker peptide, in Synechocystis sp. PCC 6803, a cyanobacterium

Phycobilisome (PBS) is a large pigment-protein complex in cyanobacteria and red algae responsible for capturing sunlight and transferring its energy to photosystems (PS). Spectroscopic and structural properties of various PBSs have been widely studied, however, the nature of so-called complex-complex interactions between PBS and PSs remains much less explored. In this work, we have investigated the function of a newly identified PBS linker protein, ApcG, some domain of which, together with a loop region (PB-loop in ApcE), is possibly located near the PBS-PS interface. Using Synechocystis sp. PCC 6803, we generated an ApcG deletion mutant and probed its deletion effect on the energetic coupling between PBS and photosystems. Steady-state and time-resolved spectroscopic characterization of the purified ΔApcG-PBS demonstrated that ApcG removal weakly affects the photophysical properties of PBS that the spectroscopic properties of terminal energy emitters are comparable to those of PBS from wild-type. However, analysis of fluorescence decay imaging datasets reveals that ApcG deletion induces disruptions within the allophycocyanin (APC) core, resulting in the emergence (splitting) of two spectrally diverse subgroups with some short-lived APC. Profound spectroscopic changes of the whole ΔApcG mutant cell, however, emerge during state transition, a dynamic process of light scheme adaptation. The mutant cells in State I show a substantial increase in PBS-related fluorescence. On the other hand, global analysis of time-resolved fluorescence demonstrates that in general ApcG deletion does not alter or inhibit state transitions if it is interpreted only in terms of the changes of the PSII and PSI fluorescence emission intensity. Furthermore, the results revealed yet–to–be discovered mechanism of ApcG-docking induced excitation energy transfer regulation within PBS or to Photosystems.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

The picasso map-making code: application to a simulation of the QUIJOTE northern sky survey

ABSTRACT Map-making is an important step for the data analysis of cosmic microwave background (CMB) experiments. It consists of converting the data, which are typically a long, complex, and noisy collection of measurements, into a map, which is an image of the observed sky. We present in this paper a new map-making code named picasso (Polarization and Intensity CArtographer for Scanned Sky Observations), which was implemented to construct intensity and polarization maps from the Multi Frequency Instrument (MFI) of the QUIJOTE (Q-U-I Joint TEnerife) CMB polarization experiment. picasso is based on the destriping algorithm, and is suited to address specific issues of ground-based microwave observations, with a technique that allows the fit of a template function in the time domain, during the map-making step. This paper describes the picasso code, validating it with simulations and assessing its performance. For this purpose, we produced realistic simulations of the QUIJOTE-MFI survey of the northern sky (approximately ∼20 000 deg2), and analysed the reconstructed maps with picasso, using real and harmonic space statistics. We show that, for this sky area, picasso is able to reconstruct, with high fidelity, the injected signal, recovering all the scales with ℓ > 10 in TT, EE, and BB. The signal error is better than 0.001 per cent at 20 < ℓ < 200. Finally, we validated some of the methods that will be applied to the real wide-survey data, like the detection of the CMB anisotropies via cross-correlation analyses. Despite that the implementation of picasso is specific for QUIJOTE-MFI data, it could be adapted to other experiments.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Identification of a broadly conserved family of enzymes that hydrolyze (p)ppApp

Bacteria produce a variety of nucleotide second messengers to adapt to their surroundings. Although chemically similar, the nucleotides guanosine penta- and tetraphosphate [(p)ppGpp] and adenosine penta- and tetraphosphate [(p)ppApp] have distinct functions in bacteria. (p)ppGpp mediates survival under nutrient-limiting conditions and its intracellular levels are regulated by synthetases and hydrolases belonging to the RelA-SpoT homolog (RSH) family of enzymes. By contrast, (p)ppApp is not known to be involved in nutrient stress responses and is synthesized by RSH-resembling toxins that inhibit the growth of bacterial cells. However, it remains unclear whether there exists a family of hydrolases that specifically act on (p)ppApp to reverse its toxic effects. Here, we present the structure and biochemical characterization of adenosine 3'-pyrophosphohydrolase 1 (Aph1), the founding member of a monofunctional (p)ppApp hydrolase family of enzymes. Our work reveals that Aph1 adopts a histidine-aspartate (HD)-domain fold characteristic of phosphohydrolase metalloenzymes and its activity mitigates the growth inhibitory effects of (p)ppApp-synthesizing toxins. Using an informatic approach, we identify over 2,000 putative (p)ppApp hydrolases that are widely distributed across bacterial phyla and found in diverse genomic contexts, and we demonstrate that 12 representative members hydrolyze ppApp. In addition, our in silico analyses reveal a unique molecular signature that is specific to (p)ppApp hydrolases, and we show that mutation of two residues within this signature broadens the specificity of Aph1 to promiscuously hydrolyze (p)ppGpp in vitro. Overall, our findings indicate that like (p)ppGpp hydrolases, (p)ppApp hydrolases are widespread in bacteria and may play important and underappreciated role(s) in bacterial physiology.

(p)ppApp↗
DOE OSTI2026

F-Hash: Feature-Based Hash Design for Time-Varying Volume Visualization via Multi-Resolution Tesseract Encoding

Interactive time-varying volume visualization is challenging due to its complex spatiotemporal features and sheer size of the dataset. Recent works transform the original discrete time-varying volumetric data into continuous Implicit Neural Representations (INR) to address the issues of compression, rendering, and super-resolution in both spatial and temporal domains. However, training the INR takes a long time to converge, especially when handling large-scale time-varying volumetric datasets. In this work, we proposed F-Hash, a novel feature-based multi-resolution Tesseract encoding architecture to greatly enhance the convergence speed compared with existing input encoding methods for modeling time-varying volumetric data. The proposed design incorporates multi-level collision-free hash functions that map dynamic 4D multi-resolution embedding grids without bucket waste, achieving high encoding capacity with compact encoding parameters. Our encoding method is agnostic to time-varying feature detection methods, making it a unified encoding solution for feature tracking and evolution visualization. Experiments show the F-Hash achieves state-of-the-art convergence speed in training various time-varying volumetric datasets for diverse features. We also proposed an adaptive ray marching algorithm to optimize the sample streaming for faster rendering of the time-varying neural representation.

deep learning↗
DOE OSTI2023

Code modernization strategies for short-range non-bonded molecular dynamics simulations

Modern HPC systems are increasingly relying on greater core counts and wider vector registers. Thus, applications need to be adapted to fully utilize these hardware capabilities. One class of applications that can benefit from this increase in parallelism are molecular dynamics simulations. In this paper, we describe our efforts at modernizing the ESPResSo++ simulation package for molecular dynamics by restructuring its particle data layout for efficient memory accesses and applying vectorization techniques to benefit the calculation of short-range non-bonded forces, which results in an overall three times speedup and serves as a baseline for further optimizations. We also implement fine-grained parallelism for multi-core CPUs through HPX, a C++ runtime system which uses lightweight threads and an asynchronous many-task approach to maximize concurrency. Our goal is to evaluate the performance of an HPX-based approach compared to the bulk-synchronous MPI-based implementation. This requires the introduction of an additional layer to the domain decomposition scheme that defines the task granularity. On spatially inhomogeneous systems, which impose a corresponding load-imbalance in traditional MPI-based approaches, we demonstrate that by choosing an optimal task size, the efficient work-stealing mechanisms of HPX can overcome the overhead of communication resulting in an overall 1.4 times speedup compared to the baseline MPI version.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Reinforcement Learning for feedback-enabled cyber resilience

The rapid growth in the number of devices and their connectivity has enlarged the attack surface and made cyber systems more vulnerable. As attackers become increasingly sophisticated and resourceful, mere reliance on traditional cyber protection, such as intrusion detection, firewalls, and encryption, is insufficient to secure the cyber systems. Cyber resilience provides a new security paradigm that complements inadequate protection with resilience mechanisms. A Cyber-Resilient Mechanism (CRM) adapts to the known or zero-day threats and uncertainties in real-time and strategically responds to them to maintain the critical functions of the cyber systems in the event of successful attacks. Feedback architectures play a pivotal role in enabling the online sensing, reasoning, and actuation process of the CRM. Reinforcement Learning (RL) is an important gathering of algorithms that epitomize the feedback architectures for cyber resilience. It allows the CRM to provide dynamic and sequential responses to attacks with limited or without prior knowledge of the environment and the attacker. In this work, we review the literature on RL for cyber resilience and discuss the cyber-resilient defenses against three major types of vulnerabilities, i.e., posture-related, information-related, and human-related vulnerabilities. Here we introduce moving target defense, defensive cyber deception, and assistive human security technologies as three application domains of CRMs to elaborate on their designs. The RL algorithms also have vulnerabilities themselves. We explain the major vulnerabilities of RL and present develop several attack models where the attacker target the information exchanged between the environment and the agent: the rewards, the state observations, and the action commands. We show that the attacker can trick the RL agent into learning a nefarious policy with minimum attacking effort. The paper introduces several defense methods to secure the RL-enabled systems from these attacks. However, there is still a lack of works that focuses on the defensive mechanisms for RL-enabled systems. Last but not least, we discuss the future challenges of RL for cyber security and resilience and emerging applications of RL-based CRMs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

N-Glycolylneuraminic Acid Binding of Avian and Equine H7 Influenza A Viruses

Influenza A viruses (IAV) initiate infection by binding to glycans with terminal sialic acids on the cell surface. Hosts of IAV variably express two major forms of sialic acid, N-acetylneuraminic acid (NeuAc) and N-glycolylneuraminic acid (NeuGc). NeuGc is produced in most mammals, including horses and pigs, but is absent in humans, ferrets, and birds. The only known naturally occurring IAV that exclusively bind NeuGc are extinct highly pathogenic equine H7N7 viruses. We determined the crystal structure of a representative equine H7 hemagglutinin (HA) in complex with NeuGc and observed high similarity in the receptor-binding domain with an avian H7 HA. To determine the molecular basis for NeuAc and NeuGc specificity, we performed systematic mutational analyses, based on the structural insights, on two distant avian H7 HAs and an H15 HA. We found that the A135E mutation is key for binding α2,3-linked NeuGc but does not abolish NeuAc binding. The additional mutations S128T, I130V, T189A, and K193R converted the specificity from NeuAc to NeuGc. We investigated the residues at positions 128, 130, 135, 189, and 193 in a phylogenetic analysis of avian and equine H7 HAs. This analysis revealed a clear distinction between equine and avian residues. The highest variability was observed at key position 135, of which only the equine glutamic acid led to NeuGc binding. These results demonstrate that genetically distinct H7 and H15 HAs can be switched from NeuAc to NeuGc binding and vice versa after the introduction of several mutations, providing insights into the adaptation of H7 viruses to NeuGc receptors.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Connected Loads – Grid Connected Appliances: Deployment IoT Solution for Fault Detection and Diagnostics

As one of the most energy-intensive end-uses in the commercial buildings sector, supermarkets consume around 50 kWh/ft 2 ( or 537.6 kWh/m 2 ) of electricity annually, or more than 2 million kWh of electricity per year for a typical store. The biggest consumer of energy in a supermarket is its refrigeration system, which accounts for 40–60% of its total electricity usage and is equivalent to about 2–3% of the total energy consumed by commercial buildings in United States, or around 0.5 quadrillion Btu (or 0.53 quadrillion KJ). Also, the supermarket refrigeration system is one of the biggest consumers of refrigerants. Current supermarket refrigeration systems rely on high global warming potential hydrofluorocarbon refrigerants. Reducing refrigerant usage or using environment friendly alternatives can result in significant climate benefits. Transcritical CO2 refrigeration systems have attracted more attention in recent years because of their zero-carbon emission advantages compared with traditional refrigerants. These systems are widely used in commercial buildings such as supermarkets. The refrigeration system can also be adapted to handle flexible building loads and be integrated into grid response transactive control to balance the supply and demand of the electric grid. Even minor improvements in the efficiency and operational reliability of supermarket refrigeration systems can create significant value in terms of saving energy, improving food quality, protecting the environment, reducing carbon footprint, and enhancing electric grid resilience. For decarbonization, the new administration has set a target to reduce greenhouse gas emissions by 50– 52% by 2030 and targeting a carbon-neutral economy by 2050. For electrification, the goal is to achieve 100% clean electricity by 2035. Such decarbonization and electrification in the building sector require that energy consumption in buildings be reduced significantly. Therefore, the building sector must continuously adopt new technologies to achieve its energy and carbon emission goals. One of the most fundamental technologies is the Internet of Things (IoT). IoT has proven to be an effective solution for the building domain, including building information/energy modeling, smart buildings, etc. Although much progress has been made in the development of IoT-based building energy systems, there is still a lack of reliable, scalable, and affordable IoT-based automated fault and degradation diagnostics (AFDDs) solutions. Such solutions would enable deployment of advanced algorithms into real systems to archive the projected energy benefits. This study reviews existing IoT solutions developed for building energy– related application and developed a simple but effective AFDD IoT deployment solution, including developing a suitable IoT architecture and conducting easy and scalable deployment by leveraging a common cloud-based IoT service.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2020

Structural adaptation of oxygen tolerance in 4-hydroxybutyrl-CoA dehydratase, a key enzyme of archaeal carbon fixation

Abstract Autotrophic microorganisms that convert inorganic carbon into organic matter were key players in the evolution of life on Earth. As the early atmosphere became oxygenated, microorganisms needed to develop mechanisms for oxygen protection, especially those relying on enzymes containing oxygen-sensitive metal clusters (e.g., Fe-S). Here we investigated how 4-hydroxybutyryl-CoA dehydratase (4HBD) - the key enzyme of the 3-hydroxypropionate/4-hydroxybutyrate (HP/HB) cycle for CO 2 -fixation - adapted as conditions shifted from anoxic to oxic. 4HBD is found in both anaerobic bacteria and aerobic ammonia-oxidizing archaea (AOA). The oxygen-sensitive bacterial 4HBD and oxygen-tolerant archaeal 4HBD share 59 % amino acid identity. To examine the structural basis of oxygen tolerance in archaeal 4HBD, we determined the atomic resolution structure of the enzyme. Two tunnels providing access to the canonical [4Fe-4S] cluster in oxygen-sensitive bacterial 4HBD were closed with four conserved mutations found in all aerobic AOA and other archaea. Further biochemical experiments and molecular dynamics simulations support our findings that restricting access to the active site is the key to oxygen tolerance, explaining how active site evolution drove a major evolutionary transition. Significance statement Autotrophy (primary production) was the first life strategy on Earth. Before photosynthesis (using solar energy to fix carbon dioxide), life relied on chemical reactions for energy. These chemosynthetic reactions are present in all domains of life, including archaea possessing the most energy-efficient carbon fixation pathway - the 3-hydroxypropionate/4-hydroxybutyrate cycle. This efficiency results from enzyme modifications, including enhanced enzyme stability and catalysis of multiple reactions. We reveal the first structure of aerobic 4-hydroxybutyryl-CoA dehydratase (4HBD) from ammonia-oxidizing archaea. These archaea are among the most abundant organisms on the planet, and their 4HBD active site evolved oxygen tolerance to support aerobic metabolism. This modification can provide further insight into enzyme evolution on early earth, as photosynthesis developed and began oxygenating the atmosphere.

DeMirci, Hasan (ORCID:0000000291355397)↗
DOE OSTI2024

Thermobifida fusca Cel6B moves bidirectionally while processively degrading cellulose

Abstract Background Cellulose, an abundant biopolymer, has great potential to be utilized as a renewable fuel feedstock through its enzymatic degradation into soluble sugars followed by sugar fermentation into liquid biofuels. However, crystalline cellulose is highly resistant to hydrolysis, thus industrial-scale production of cellulosic biofuels has been cost-prohibitive to date. Mechanistic studies of enzymes that break down cellulose, called cellulases, are necessary to improve and adapt such biocatalysts for implementation in biofuel production processes. Thermobifida fusca Cel6B ( Tf Cel6B) is a promising candidate for industrial use due to its thermostability and insensitivity to pH changes. However, mechanistic studies probing Tf Cel6B hydrolytic activity have been limited to ensemble-scale measurements. Results We utilized optical tweezers to perform single-molecule, nanometer-scale measurements of enzyme displacement during cellulose hydrolysis by Tf Cel6B. Records featured forward motility on the order of 0.17 nm s −1 interrupted by backward motions and long pauses. Processive run lengths were on the order of 5 nm in both forward and backward directions. Motility records also showed rapid bidirectional displacements greater than 5 nm. Single-enzyme velocity and bulk ensemble activity were assayed on multiple crystalline cellulose allomorphs revealing that the degree of crystallinity and hydrogen bonding have disparate effects on the single-molecule level compared to the bulk scale. Additionally, we isolated and monitored the catalytic domain of Tf Cel6B and observed a reduction in velocity compared to the full-length enzyme that includes the carbohydrate-binding module. Applied force has little impact on enzyme velocity yet it readily facilitates dissociation from cellulose. Preliminary measurements at elevated temperatures indicated enzyme velocity strongly increases with temperature. Conclusions The unexpected motility patterns of Tf Cel6B are likely due to previously unknown mechanisms of processive cellulase motility implicating irregularities in cellulose substrate ultrastructure. While Tf Cel6B is processive, it has low motility at room temperature. Factors that most dramatically impact enzyme velocity are temperature and the presence of its native carbohydrate-binding module and linker. In contrast, substrate ultrastructure and applied force did not greatly impact velocity. These findings motivate further study of Tf Cel6B for its engineering and potential implementation in industrial processes.

Johnson, Madeline M. (ORCID:000000029518293X)↗
DOE OSTI2023

Automating methods for estimating metabolite volatility

The volatility of metabolites can influence their biological roles and inform optimal methods for their detection. Yet, volatility information is not readily available for the large number of described metabolites, limiting the exploration of volatility as a fundamental trait of metabolites. Here, we adapted methods to estimate vapor pressure from the functional group composition of individual molecules (SIMPOL.1) to predict the gas-phase partitioning of compounds in different environments. We implemented these methods in a new open pipeline called volcalc that uses chemoinformatic tools to automate these volatility estimates for all metabolites in an extensive and continuously updated pathway database: the Kyoto Encyclopedia of Genes and Genomes (KEGG) that connects metabolites, organisms, and reactions. We first benchmark the automated pipeline against a manually curated data set and show that the same category of volatility (e.g., nonvolatile, low, moderate, high) is predicted for 93% of compounds. We then demonstrate how volcalc might be used to generate and test hypotheses about the role of volatility in biological systems and organisms. Specifically, we estimate that 3.4 and 26.6% of compounds in KEGG have high volatility depending on the environment (soil vs. clean atmosphere, respectively) and that a core set of volatiles is shared among all domains of life (30%) with the largest proportion of kingdom-specific volatiles identified in bacteria. With volcalc , we lay a foundation for uncovering the role of the volatilome using an approach that is easily integrated with other bioinformatic pipelines and can be continually refined to consider additional dimensions to volatility. The volcalc package is an accessible tool to help design and test hypotheses on volatile metabolites and their unique roles in biological systems.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Redox regulation of m 6 A methyltransferase METTL3 in β-cells controls the innate immune response in type 1 diabetes

Type 1 Diabetes (T1D) is characterized by the destruction of pancreatic ß-cells. Several observations have renewed the interest in ß-cell RNA sensors and editors. Here, in this study, we report that N 6 -methyladenosine (m 6 A) is an adaptive ß-cell safeguard mechanism that controls the amplitude and duration of the antiviral innate immune response at T1D onset. m 6 A writer methyltransferase 3 (METTL3) levels increase drastically in ß-cells at T1D onset but rapidly decline with disease progression. m 6 A-sequencing revealed the m 6 A hypermethylation of several key innate immune mediators including OAS1, OAS2, OAS3, and ADAR1 in human islets and EndoC-ßH1 cells at T1D onset. METTL3 silencing enhanced OAS levels by increasing its mRNA stability. Consistently, in vivo gene therapy, to prolong Mettl3 overexpression specifically in ß-cells, delayed diabetes progression in the non-obese diabetic (NOD) mouse model of T1D. Mechanistically, the accumulation of reactive oxygen species blocked upregulation of METTL3 in response to cytokines, while physiological levels of nitric oxide enhanced its expression. Furthermore, for the first time to our knowledge, we report that the cysteines in position C276 and C326 in the zinc finger domains of the METTL3 protein are sensitive to S-nitrosylation (SNO) and are significant for the METTL3-mediated regulation of OAS mRNA stability in human ß-cells. Collectively, we report that m 6 A regulates ß-cells to control the innate immune response during the onset of T1D in humans.

2'-5'-oligoadenylate synthetase↗
DOE OSTI2022

(Towards) DNS of a Laboratory Lean CH4/H2 Low-Swirl Flame Impinging on an Inclined Wall

Due to downsizing trends, flame-wall interaction (FWI) is increasingly prominent in gas turbines (GTs). FWI has direct consequences on flame stabilization and pollutant emissions, but it is not well understood in turbulent flows representative of GTs. We present results from a direct numerical simulation (DNS) of a turbulent CH4/H2 model GT low-swirl laboratory-scale flame interacting with an inclined wall. The results from the laboratory flame include simultaneous measurements of velocity using stereo particle imaging velocimetry and OHxCH2O planar laser induced fluorescence. The adaptive-mesh refinement solver PeleLMeX is used, with 24-species, 105-reaction reduced Aramco chemical kinetics mechanism. The premixed fuel-air mixture consists of hydrogen-enriched methane with 70% hydrogen volume fraction and 0.4 equivalence ratio. The inflow is prescribed to match experimental measurements at the burner exit. Karlovitz and turbulent Reynolds numbers are 300 and 400, respectively. The simulation and experimental results show excellent agreement. The flame features a bowl-shape stabilization, with a corrugated, continuous flame front at the leading edge, followed by fragmented reaction zones downstream. A large diffuse cloud of CH2O is formed downstream of the quenching point. The simulation results indicate that the cloud of CH2O is the result of incomplete methane combustion, with CH2O "leaking" from the locally quenched reaction zones.The DNS provides fine-grain resolution of turbulence-flame-wall interaction that cannot be captured with experimental measurements. With access to the entire solution vector at each cell of the computational domain, the local quenching.

flame-wall interactions↗
DOE OSTI2026

Queue wait time prediction in high performance computing (HPC) systems

High Performance Computing (HPC) systems are critical enablers for groundbreaking scientific research across various domains. Efficient resource allocation, facilitated by job scheduling, is paramount for maximizing the utilization of HPC systems. However, the variability in wait times for queued jobs poses challenges for users, necessitating accurate job wait time estimation. This paper explores the influence of job characteristics, including job size (the number of nodes requested and walltime), the queue to which the job is submitted and other resource requirements, on job wait times in leadership-class HPC systems. Focusing on the Theta Cray XC40 and Polaris machines at Argonne National Laboratory, the study evaluates the performance of different supervised learning algorithms in predicting job wait times. It also evaluates the impact of data preprocessing, including outlier detection, Principal Component Analysis (PCA), and feature selection, on the performance of wait time prediction models. The findings reveal insights into the relationship between job characteristics and wait times, offering a foundation for optimizing resource allocation and enhancing user experience. The methodologies and tools developed in this study are adaptable to other leadership-class HPC systems, providing a valuable contribution to the broader HPC community aiming to improve job scheduling efficiency and user satisfaction.

Okafor, Nwamaka↗
DOE OSTI2024

CG-Kit: Code Generation Toolkit for performant and maintainable variants of source code applied to Flash-X hydrodynamics simulations

CG-Kit is a new Code Generation tool-Kit that we have developed as a part of the solution for portability and maintainability for multiphysics computing applications. The development of CG-Kit is rooted in the urgent need created by the shifting landscape of high-performance computing platforms and the algorithmic complexities of a particular large-scale multiphysics application: Flash-X. To efficiently use computing resources on a heterogeneous node, an application must have a map of computation to resources and a mechanism to move the data and computation to the resources according to the map. Most existing performance portability solutions are focussed on abstracting the expression of computations so that a unified source code can be specialized to run on different resources. However, such an approach is insufficient for a code like Flash-X, which has a multitude of code components that can be assembled in various permutations and combinations to form different instances of applications. Similar challenges apply to any code that has composability, where a single specified way of apportioning work among devices may not be optimal. Additionally, use cases arise where the optimal control flow of computation may differ for different devices while the underlying numerics remain identical. This combination leads to unique challenges including handling an existing large code base in Fortran and/or C/C++, subdivision of code into a great variety of units supporting a wide range of physics and numerical methods, different parallelization techniques for distributed and shared memory systems and accelerator devices, and heterogeneity of computing platforms requiring coexisting variants of parallel algorithms. All of these challenges demand that scientific software developers apply existing knowledge about domain applications, algorithms, and computing platforms to determine custom abstractions and granularity for code generation. There is a critical lack of tools to tackle those problems. CG-Kit is designed to fill this gap by providing a user with the ability to express their desired control flow and computation-to-resource map in the form a pseudocode-like recipe. It consists of standalone tools that can be combined into highly specific and, we argue, highly effective portability and maintainability toolchains. Here we present the design of our new tools: parametrized source trees, control flow graphs, and recipes. The tools are implemented in Python. They are agnostic to the programming language of the source code targeted for code generation. In conclusion, we demonstrate the capabilities of the toolkit with two examples, first, multithreaded variants of the basic AXPY operation, and second, variants of parallel algorithms within a hydrodynamics solver, called Spark, from Flash-X that operates on block-structured adaptive meshes.

Algorithmic portability↗
DOE OSTI2021

Comparison of the Forward and Reverse Photocycle Dynamics of Two Highly Similar Canonical Red/Green Cyanobacteriochromes Reveals Unexpected Differences

Cyanobacteriochromes (CBCRs) are cyanobacterial photoreceptors that exhibit photochromism between two states: a thermally stable dark-adapted state and a metastable light-adapted state with bound linear tetrapyrrole (bilin) chromophores possessing 15Z and 15E configurations, respectively. The photodynamics of canonical red/green CBCRs have been extensively studied; however, the time scales of their excited-state lifetimes and subsequent ground-state evolution rates widely differ and, at present, remain difficult to predict. Here, we compare the photodynamics of two closely related red/green CBCRs that have substantial sequence identity (~68%) and similar chromophore environments: AnPixJg2 from Anabaena sp. PCC 7120 and NpR6012g4 from Nostoc punctiforme. Using broadband transient absorption spectroscopy on the primary (125 fs to 7 ns) and secondary (7 ns to 10 ms) time scales together with global analysis modeling, our studies revealed that AnPixJg2 and NpR6012g4 have comparable quantum yields for initiating the forward ( 15Z P r → 15E P g ) and reverse ( 15E P g → 15Z P r ) reactions, which proceed through monotonic and nonmonotonic mechanisms, respectively. In addition to small discrepancies in the kinetics, the secondary reverse dynamics resolved unique features for each domain: intermediate shunts in NpR6012g4 and a Meta-Gf intermediate red-shifted from the 15Z P r photoproduct in AnPixJg2. Altogether, this study supports the conclusion that sequence similarity is a useful criterion for predicting pathways of the light-induced evolution and quantum yield of generating primary intermediate Φ p within subfamilies of CBCRs, but more studies are still needed to develop a comprehensive molecular level understanding of these processes.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

OmniXAS: A universal deep-learning framework for materials x-ray absorption spectra

X-ray absorption spectroscopy (XAS) is a powerful characterization technique for probing the local chemical environment of absorbing atoms. However, analyzing XAS data presents significant challenges, often requiring extensive, computationally intensive simulations, as well as significant domain expertise. These limitations hinder the development of fast, robust XAS analysis pipelines that are essential in high-throughput studies and for autonomous experimentation. Here, we address these challenges with OmniXAS, a framework that contains a suite of transfer learning approaches for XAS prediction, each uniquely contributing to improved accuracy and efficiency, as demonstrated on the K-edge spectra database covering eight 3⁢d transition metals (Ti–Cu). The OmniXAS framework is built upon three distinct strategies. First, we use M3GNet [Nat. Comput. Sci. 2, 718 (2022)] to derive latent representations of the local chemical environment of absorption sites as input for XAS prediction, achieving significant improvements over conventional featurization techniques. Second, we employ a hierarchical transfer learning strategy, training a universal multitask model across elements before fine-tuning for element-specific predictions. Models based on this cascaded approach after elementwise fine-tuning outperform element-specific models by up to 69%. Third, we implement cross-fidelity transfer learning, adapting a universal model to predict spectra generated by simulation of a different fidelity with a much higher computational cost. This approach improves prediction accuracy by up to 11% over models trained on the target fidelity alone. Our approach significantly boosts the throughput of XAS modeling by orders of magnitude as compared to first-principles simulations and is extendable to XAS prediction for a broader range of elements. The proposed transfer learning framework is generalizable to enhance deep-learning models that target other properties in materials research.

36 MATERIALS SCIENCE↗