Evaluating point-prediction uncertainties in neural networks for protein-ligand binding prediction
Not Available
97 MATHEMATICS AND COMPUTING↗
SEARCH · Search NASA
Results for “domain of applicability”
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.
Local Decomposition of Hexahedral Singular Nodes into Singular Curves
Hexahedral (hex) meshing is a long studied topic in geometry processing with many challenging associated problems. Hex meshes vary from structured to unstructured depending on application or domain of interest. Fully structured meshes require that all interior mesh edges be adjacent to four hexes each. Edges failing this criteria are singular and indicate an unstructured hex mesh. Singular edges join together into singular curves that either form closed cycles, end on the mesh boundary, or end at a singular node, a complex junction of more than two singular curves. Hex meshes with more complex singular nodes tend to have more distorted elements and smaller scaled Jacobian values. In this work, we study the topology of singular nodes. We show that all eight of the most common singular nodes are decomposable into just singular curves. We further show that all singular nodes, regardless of edge valence, are locally decomposable. Finally we demonstrate these decompositions on hex meshes, thereby decreasing their distortion and converting all singular nodes into singular curves. In conclusion, with this decomposition, the enigmatic complexity of 3D singular nodes becomes effectively 2D.
97 MATHEMATICS AND COMPUTING↗
Solid face sheets enable lattice metamaterials to withstand high-amplitude impulsive loading without yielding
Owing to their ability to provide tunable mechanical responses, lattice materials are frequently studied to elucidate their response to static and dynamic loads. However, these roles are typically in opposition: static loads must be supported sufficiently far away from the onset of buckling or yielding, whereas dynamic loads are typically ameliorated by crushing of the lattice, which provides excellent energy-absorption due to the large plastic deformation accompanying densification. In contrast, this work considers the octet truss as an exemplar topology, in a structural role where it must simultaneously support static loads while enduring high-amplitude impulsive loads. This study focuses on the ability to withstand impulsive loads without yielding, an essential prerequisite to enduring dual loading. Computational studies using the ALE3D hydrocode were performed to examine the response of the octet truss under a short temporal width impulse shape associated with laser-driven shocks. A key finding was that covering the lattice with a solid face sheet and treating this face sheet thickness as a design variable allows the Taylor-like pulse to be attenuated prior to entering the weaker lattice, at the cost of added mass up front. Experimental validation was accomplished by laser-driven shock testing, using octet trusses printed out of Ti-5Al-5V-5Mo-3Cr. The results show that for a given quantity of mass, the attenuation is maximized when as much mass as possible is moved into the face sheet, leaving a more slender lattice structure. The effect of placing mass in the face sheet rather than lattice beams dominates the effect of relative density, to the point where a low-mass structure with most of the mass concentrated in the face sheet can outperform a high-mass structure with most of the mass in the lattice. Finally, by further understanding the propagation of short pulse width waves within under-dense structures, this study expand the domain of applicability of such structures, including lattice materials, to challenging dual-loading regimes spanning decades of strain rates.
36 MATERIALS SCIENCE↗
Navigating Transition-Metal Chemical Space: Artificial Intelligence for First-Principles Design
Conspectus The variability of chemical bonding in open-shell transition-metal complexes not only motivates their study as functional materials and catalysts but also challenges conventional computational modeling tools. Here, tailoring ligand chemistry can alter preferred spin or oxidation states as well as electronic structure properties and reactivity, creating vast regions of chemical space to explore when designing new materials atom by atom. Although first-principles density functional theory (DFT) remains the workhorse of computational chemistry in mechanism deduction and property prediction, it is of limited use here. DFT is both far too computationally costly for widespread exploration of transition-metal chemical space and also prone to inaccuracies that limit its predictive performance for localized d electrons in transition-metal complexes. These challenges starkly contrast with the well-trodden regions of small-organic-molecule chemical space, where the analytical forms of molecular mechanics force fields and semiempirical theories have for decades accelerated the discovery of new molecules, accurate DFT functional performance has been demonstrated, and gold-standard methods from correlated wavefunction theory can predict experimental results to chemical accuracy. The combined promise of transition-metal chemical space exploration and lack of established tools has mandated a distinct approach. In this Account, we outline the path we charted in exploration of transition-metal chemical space starting from the first machine learning (ML) models (i.e., artificial neural network and kernel ridge regression) and representations for the prediction of open-shell transition-metal complex properties. The distinct importance of the immediate coordination environment of the metal center as well as the lack of low-level methods to accurately predict structural properties in this coordination environment first motivated and then benefited from these ML models and representations. Once developed, the recipe for prediction of geometric, spin state, and redox potential properties was straightforwardly extended to a diverse range of other properties, including in catalysis, computational “feasibility”, and the gas separation properties of periodic metal–organic frameworks. Interpretation of selected features most important for model prediction revealed new ways to encapsulate design rules and confirmed that models were robustly mapping essential structure–property relationships. Encountering the special challenge of ensuring that good model performance could generalize to new discovery targets motivated investigation of how to best carry out model uncertainty quantification. Distance-based approaches, whether in model latent space or in carefully engineered feature space, provided intuitive measures of the domain of applicability. With all of these pieces together, ML can be harnessed as an engine to tackle the large-scale exploration of transition-metal chemical space needed to satisfy multiple objectives using efficient global optimization methods. In practical terms, bringing these artificial intelligence tools to bear on the problems of transition-metal chemical space exploration has resulted in ML-model assessments of large, multimillion compound spaces in minutes and validated new design leads in weeks instead of decades.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Predicting Small Molecule Transfer Free Energies by Combining Molecular Dynamics Simulations and Deep Learning
Accurately predicting small molecule partitioning and hydrophobicity is critical in the drug discovery process. There are many heterogeneous chemical environments within a cell and entire human body. For example, drugs must be able to cross the hydrophobic cellular membrane to reach their intracellular targets, and hydrophobicity is an important driving force for drug–protein binding. Atomistic molecular dynamics (MD) simulations are routinely used to calculate free energies of small molecules binding to proteins, crossing lipid membranes, and solvation but are computationally expensive. Machine learning (ML) and empirical methods are also used throughout drug discovery but rely on experimental data, limiting the domain of applicability. We present atomistic MD simulations calculating 15,000 small molecule free energies of transfer from water to cyclohexane. This large data set is used to train ML models that predict the free energies of transfer. We show that a spatial graph neural network model achieves the highest accuracy, followed closely by a 3D-convolutional neural network, and shallow learning based on the chemical fingerprint is significantly less accurate. A mean absolute error of ~4 kJ/mol compared to the MD calculations was achieved for our best ML model. We also show that including data from the MD simulation improves the predictions, tests the transferability of each model to a diverse set of molecules, and show multitask learning improves the predictions. This work provides insight into the hydrophobicity of small molecules and ML cheminformatics modeling, and our data set will be useful for designing and testing future ML cheminformatics methods.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Graph-Based Approaches for Predicting Solvation Energy in Multiple Solvents: Open Datasets and Machine Learning Models
The solvation properties of molecules, often estimated using quantum chemical simulations, are important in the synthesis of energy storage materials, drugs, and industrial chemicals. Here, we develop machine learning models of solvation energies to replace expensive quantum chemistry calculations with inexpensive-to-compute message-passing neural network models that require only the molecular graph as inputs. Our models are trained on a new database of solvation energies for 130,258 molecules taken from the QM9 dataset computed in five solvents (acetone, ethanol, acetonitrile, dimethyl sulfoxide, and water) via an implicit solvent model. Our best model achieves a mean absolute error of 0.5 kcal/mol for molecules with nine or fewer non-hydrogen atoms and 1 kcal/mol for molecules with between 10 and 14 non-hydrogen atoms. We make the entire dataset of 651,290 computed entries openly available and provide simple web and programmatic interfaces to enable others to run our solvation energy model on new molecules. This model calculates the solvation energies for molecules using only the SMILES string and also provides an estimate of whether each molecule is within the domain of applicability of our model. We envision that the dataset and models will provide the functionality needed for the rapid screening of large chemical spaces to discover improved molecules for many applications.
25 ENERGY STORAGE↗
The emergence of macroscopic currents in photoconductive sampling of optical fields
Photoconductive field sampling enables petahertz-domain optoelectronic applications that advance our understanding of light-matter interaction. Despite the growing importance of ultrafast photoconductive measurements, a rigorous model for connecting the microscopic electron dynamics to the macroscopic external signal is lacking. This has caused conflicting interpretations about the origin of macroscopic currents. Here, we present systematic experimental studies on the signal formation in gas-phase photoconductive sampling. Our theoretical model, based on the Ramo–Shockley-theorem, overcomes the previously introduced artificial separation into dipole and current contributions. Extensive numerical particle-in-cell-type simulations permit a quantitative comparison with experimental results and help to identify the roles of electron-neutral scattering and mean-field charge interactions. The results show that the heuristic models utilized so far are valid only in a limited range and are affected by macroscopic effects. Our approach can aid in the design of more sensitive and more efficient photoconductive devices.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Uncertainty quantification for molecular property predictions with graph neural architecture search
Graph Neural Networks (GNNs) have emerged as a prominent class of data-driven methods for molecular property prediction. However, a key limitation of typical GNN models is their inability to quantify uncertainties in the predictions. This capability is crucial for ensuring the trustworthy use and deployment of models in downstream tasks. To that end, we introduce AutoGNNUQ, an automated uncertainty quantification (UQ) approach for molecular property prediction. AutoGNNUQ leverages architecture search to generate an ensemble of high-performing GNNs, enabling the estimation of predictive uncertainties. Our approach employs variance decomposition to separate data (aleatoric) and model (epistemic) uncertainties, providing valuable insights for reducing them. In our computational experiments, we demonstrate that AutoGNNUQ outperforms existing UQ methods in terms of both prediction accuracy and UQ performance on multiple benchmark datasets, and generalizes well to out-of-distribution datasets. Additionally, we utilize t-SNE visualization to explore correlations between molecular features and uncertainty, offering insight for dataset improvement. AutoGNNUQ has broad applicability in domains such as drug discovery and materials science, where accurate uncertainty quantification is crucial for decision-making.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
TAMM: Tensor algebra for many-body methods
Tensor algebra operations such as contractions in computational chemistry consume a significant fraction of the computing time on large-scale computing platforms. The widespread use of tensor contractions between large multi-dimensional tensors in describing electronic structure theory has motivated the development of multiple tensor algebra frameworks targeting heterogeneous computing platforms. In this paper, we present Tensor Algebra for Many-body Methods (TAMM), a framework for productive and performance-portable development of scalable computational chemistry methods. TAMM decouples the specification of the computation from the execution of these operations on available high-performance computing systems. With this design choice, the scientific application developers (domain scientists) can focus on the algorithmic requirements using the tensor algebra interface provided by TAMM, whereas high-performance computing developers can direct their attention to various optimizations on the underlying constructs, such as efficient data distribution, optimized scheduling algorithms, and efficient use of intra-node resources (e.g., graphics processing units). The modular structure of TAMM allows it to support different hardware architectures and incorporate new algorithmic advances. We describe the TAMM framework and our approach to the sustainable development of scalable ground- and excited-state electronic structure methods. We present case studies highlighting the ease of use, including the performance and productivity gains compared to other frameworks.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Impact of three-body forces on elastic nucleon-nucleus scattering observables
In a previous series of papers we investigated the domain of applicability of chiral potentials to the construction of a microscopic optical potential (OP) for elastic nucleon-nucleus scattering. The OP was derived at the first order of the spectator expansion of the Watson multiple scattering theory and its final expression was a folding integral between the nucleon-nucleon ( N N ) t matrix and the nuclear density of the target. In the calculations N N and three-nucleon ( 3 N ) chiral interactions were used for the target density and only the N N interaction for the N N t matrix. The purpose of this study is to achieve another step towards the calculation of a more consistent OP introducing the 3 N force also in the dynamic part of the OP. The full treatment of the 3 N interaction is beyond our present capabilities. Thus, here it is approximated with a density dependent N N interaction obtained after the averaging over the Fermi sphere. In practice, in our model the 3 N force acts as a medium correction of the bare N N interaction used to calculate the t matrix. Even if the 3 N force is treated in an approximate way, this method naturally extends our previous model of the OP and allows a direct comparison of our present and previous results. We consider as case studies the elastic scattering of nucleons off C 12 and O 16 . We present results for the differential cross section and the spin observables for different values of the projectile energy. From the comparison with the experimental data and with the results of our previous model we assess the importance of the 3 N interaction in the dynamic part of the OP. Conclusions: Our analysis indicates that the contribution of the 3 N force in the t matrix is small for the differential cross section and it is sizable for the spin observables, in particular, for the analyzing power. We find that the two-pion exchange term is the major contributor to the 3 N force. A chiral expansion order-by-order analysis of the scattering observables confirms the convergence of our results at the next-to-next-to-next-to-leading-order, as already established in our previous work.
73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
Validity of path thermodynamic description of reactive systems: Microscopic simulations
Traditional stochastic modeling of reactive systems limits the domain of applicability of the associated path thermodynamics to systems involving a single elementary reaction at the origin of each observed change in composition. An alternative stochastic modeling has recently been proposed to overcome this limitation. These two ways of modeling reactive systems are in principle incompatible. Here, the question thus arises about choosing the appropriate type of modeling to be used in practical situations. In the absence of sufficiently accurate experimental results, one way to address this issue is through the microscopic simulation of reactive fluids, usually based on hard-sphere dynamics in the Boltzmann limit. In this paper, we show that results obtained through such simulations unambiguously confirm the predictions of traditional stochastic modeling, invalidating a recently proposed alternative.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Leveraging Compiler-Based Translation to Evaluate a Diversity of Exascale Platforms
Accelerator-based heterogeneous computing is the de facto standard in current and upcoming exascale machines. These heterogeneous resources empower computational scientists to select a machine or platform well-suited to their domain or applications. However, this diversity of machines also poses challenges related to programming model selection: inconsistent availability of programming models across different exascale systems, lack of performance portability for those programming models that do span several systems, and inconsistent performance between different models on a single platform. We explore these challenges on exascale-similar hardware, including AMD MI100 and NVIDIA A100 GPUs. By extending the sourceto-source compiler OpenARC, we demonstrate the power of automated translation of applications written in a single frontend programming model (OpenACC) into a variety of backend models (OpenMP, OpenCL, CUDA, HIP) that span the upcoming exascale environments. This translation enables us to compare performance within and across devices and to analyze programming model behavior with profiling tools.
Lambert, Jacob↗
Significant DBSCAN+: Statistically Robust Density-based Clustering
Cluster detection is important and widely used in a variety of applications, including public health, public safety, transportation, and so on. Given a collection of data points, we aim to detect density-connected spatial clusters with varying geometric shapes and densities, under the constraint that the clusters are statistically significant. The problem is challenging, because many societal applications and domain science studies have low tolerance for spurious results, and clusters may have arbitrary shapes and varying densities. As a classical topic in data mining and learning, a myriad of techniques have been developed to detect clusters with both varying shapes and densities (e.g., density-based, hierarchical, spectral, or deep clustering methods). However, the vast majority of these techniques do not consider statistical rigor and are susceptible to detecting spurious clusters formed as a result of natural randomness. On the other hand, scan statistic approaches explicitly control the rate of spurious results, but they typically assume a single “hotspot” of over-density and many rely on further assumptions such as a tessellated input space. To unite the strengths of both lines of work, we propose a statistically robust formulation of a multi-scale DBSCAN, namely Significant DBSCAN+, to identify significant clusters that are density connected. As we will show, incorporation of statistical rigor is a powerful mechanism that allows the new Significant DBSCAN+ to outperform state-of-the-art clustering techniques in various scenarios. We also propose computational enhancements to speed-up the proposed approach. Experiment results show that Significant DBSCAN+ can simultaneously improve the success rate of true cluster detection (e.g., 10–20% increases in absolute F1 scores) and substantially reduce the rate of spurious results (e.g., from thousands/hundreds of spurious detections to none or just a few across 100 datasets), and the acceleration methods can improve the efficiency for both clustered and non-clustered data.
Computer Science↗
From Rules to Reasoning: A Survey of Large Language Model-Based Approaches to Scientific Hypothesis and Idea Generation
Scientific hypothesis generation represents a fundamental challenge in contemporary research due to exponentially expanding literature volumes and increasing disciplinary specialization. Large language models (LLMs) have emerged as transformative tools for automated scientific discovery, moving beyond traditional rule-based and literature-mining approaches. Four paradigmatic approaches define current LLM-driven hypothesis generation: direct prompting and fine-tuning methods, knowledge-enhanced frameworks integrating retrieval-augmented generation (RAG), multi-agent collaborative systems simulating research teams, and reasoning-focused approaches implementing cognitive architectures. Domain-specific applications demonstrate statistical equivalence to human expert performance in social psychology, experimental validation in biomedical research, and near-expert quality in astronomy. Evaluation methodologies encompass human expert assessment, LLM-as-judge frameworks, and comprehensive benchmarking systems. Technical challenges include hallucination management, knowledge integration limitations, and balancing novelty with feasibility. Future directions emphasize hybrid neural-symbolic architectures and sophisticated human-AI collaboration models for responsible scientific discovery acceleration.
AI-driven discovery↗
Java Software Extensibility Library
The Java Software Extensibility Library provides tools that help Java software developers produce flexible and extensible Java applications. The major components include a Plugin Library and a Polymorphic Map data structure. The Plugin Library provides developers with a Plugin Registry and a standard pattern for designing, versioning, registering, retrieving, and annotating plugins. It also enables advanced capabilities such as loading plugins at runtime, and creating/loading encrypted plugin packages. The Polymorphic Map data structure facilitates the storage and retrieval of objects of varying types within a single data structure; while maintaining type safety. The library also provides a standard design pattern for producing application-specific "Domain Maps" and "Closed Domain Maps".SAND2020-3816 M Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.
Fleschute, Christopher↗
pnnl/doa
This is a set of Python scripts designed to implement a method of determining the domain of applicability for a trained machine model for molecular property prediction. Additionally, the scripts implement analysis methods to evaluate and explore the results of the method. These scripts are designed to support the submission of a manuscript describing the methods for publication.
Panapitiya, Gihan↗
Orchestration of materials science workflows for heterogeneous resources at large scale
In the era of big data, materials science workflows need to handle large-scale data distribution, storage, and computation. Any of these areas can become a performance bottleneck. We present a framework for analyzing internal material structures (e.g., cracks) to mitigate these bottlenecks. We demonstrate the effectiveness of our framework for a workflow performing synchrotron X-ray computed tomography reconstruction and segmentation of a silica-based structure. Our framework provides a cloud-based, cutting-edge solution to challenges such as growing intermediate and output data and heavy resource demands during image reconstruction and segmentation. Specifically, our framework efficiently manages data storage, scaling up compute resources on the cloud. The multi-layer software structure of our framework includes three layers. A top layer uses Jupyter notebooks and serves as the user interface. A middle layer uses Ansible for resource deployment and managing the execution environment. A low layer is dedicated to resource management and provides resource management and job scheduling on heterogeneous nodes (i.e., GPU and CPU). At the core of this layer, Kubernetes supports resource management, and Dask enables large-scale job scheduling for heterogeneous resources. The broader impact of our work is four-fold: through our framework, we hide the complexity of the cloud’s software stack to the user who otherwise is required to have expertise in cloud technologies; we manage job scheduling efficiently and in a scalable manner; we enable resource elasticity and workflow orchestration at a large scale; and we facilitate moving the study of nonporous structures, which has wide applications in engineering and scientific fields, to the cloud. While we demonstrate the capability of our framework for a specific materials science application, it can be adapted for other applications and domains because of its modular, multi-layer architecture.
97 MATHEMATICS AND COMPUTING↗
Analysis of picosecond coherent anti-Stokes Raman spectra for gas-phase diagnostics
We present a hybrid frequency- and time-domain solution, applicable to the case of picosecond coherent anti-Stokes Raman scattering (CARS), for gas-phase diagnostics. A solution has been derived based on both physical arguments and four-wave mixing equations for picosecond CARS, with pulse durations that are comparable to the dephasing time scale for gas-phase Raman coherence—a regime where commonly employed solutions for impulsive (femtosecond) or cw (nanosecond) pump/Stokes forcing are not strictly valid. We present the ps-CARS spectrum in the form of incoherent sums of CARS intensity spectra, calculated from the fundamental solution for impulsive pump/Stokes Raman preparation. The solution was examined for temperatures from 1000–3000 K, for four plausible experimental configurations, with laser pulse durations of 50–150 ps, and probe pulse delays from −20 to 240 ps. Approximations based on cw and impulsive pump/Stokes preparation to fit picosecond CARS spectra at atmospheric pressure were examined and the relative thermometric accuracy and computational cost of these approximations were quantified for the case of a zero nonresonant CARS contribution, and a nonresonant susceptibility equal to 10% of the Raman-resonant value at the N 2 bandhead. The nanosecond CARS approximation can result in large fitting errors when the probe pulse time delay is less than the probe pulse duration. Errors as large as 10–20% are observed in the fit temperatures for a zero picosecond probe pulse delay, when the nonresonant background is neglected, largely due to an inability of the time-independent cw model to capture transient frequency spread dephasing effects at the Q -branch bandhead. The inclusion of a nonresonant background results in 40–60% thermometry errors with a nanosecond model at a zero-probe delay. Time-dependent impulsive calculations used for femtosecond CARS better approximate the structure of the N 2 bandhead, reducing temperature fitting errors to 5–10% at a short probe pulse delay. The impulsive approximation results in errors up to 10% at intermediate probe pulse delays, where the coherence of the pump and probe pulses leads to multiple terms in the picosecond CARS solution. Both approximations improve as the probe pulse delay exceeds the probe duration. The nanosecond approximation results in a 2–3% error, while the impulsive model results in differences of less than 1% in some cases. Fits to experimental data obtained using short, ∼60ps pulses at a zero probe time delay and longer 100 ps pulses at a substantial 200 ps delay are presented with accuracies of 1–3% in the fit temperature.
Kearney, Sean P.↗