Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “predictability”

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 1,243 records · Page 69
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Clarifying trust of materials property predictions using neural networks with distribution-specific uncertainty quantification

It is critical that machine learning (ML) model predictions be trustworthy for high-throughput catalyst discovery approaches. Uncertainty quantification (UQ) methods allow estimation of the trustworthiness of an ML model, but these methods have not been well explored in the field of heterogeneous catalysis. Herein, we investigate different UQ methods applied to a crystal graph convolutional neural network to predict adsorption energies of molecules on alloys from the Open Catalyst 2020 dataset, the largest existing heterogeneous catalyst dataset. We apply three UQ methods to the adsorption energy predictions, namely k-fold ensembling, Monte Carlo dropout, and evidential regression. The effectiveness of each UQ method is assessed based on accuracy, sharpness, dispersion, calibration, and tightness. Evidential regression is demonstrated to be a powerful approach for rapidly obtaining tunable, competitively trustworthy UQ estimates for heterogeneous catalysis applications when using neural networks. Recalibration of model uncertainties is shown to be essential in practical screening applications of catalysts using uncertainties.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Improving materials property predictions for graph neural networks with minimal feature engineering *

Graph neural networks (GNNs) have been employed in materials research to predict physical and functional properties, and have achieved superior performance in several application domains over prior machine learning approaches. Recent studies incorporate features of increasing complexity such as Gaussian radial functions, plane wave functions, and angular terms to augment the neural network models, with the expectation that these features are critical for achieving a high performance. Here, we propose a GNN that adopts edge convolution where hidden edge features evolve during training and extensive attention mechanisms, and operates on simple graphs with atoms as nodes and distances between them as edges. As a result, the same model can be used for very different tasks as no other domain-specific features are used. With a model that uses no feature engineering, we achieve performance comparable with state-of-the-art models with elaborate features for formation energy and band gap prediction with standard benchmarks; we achieve even better performance when the dataset size increases. Although some domain-specific datasets still require hand-crafted features to achieve state-of-the-art results, our selected architecture choices greatly reduce the need for elaborate feature engineering and still maintain predictive power in comparison.

42 ENGINEERING↗
DOE OSTI2024

Transferring predictions of formation energy across lattices of increasing size*

In this study, we show the transferability of graph convolutional neural network (GCNN) predictions of the formation energy of the nickel-platinum solid solution alloy across atomic structures of increasing sizes. The original dataset was generated with the large-scale atomic/molecular massively parallel simulator using the second nearest-neighbor modified embedded-atom method empirical interatomic potential. Geometry optimization was performed on the initially randomly generated face centered cubic crystal structures and the formation energy has been calculated at each step of the geometry optimization, with configurations spanning the whole compositional range. Using data from various steps of the geometry optimization, we first trained our open-source, scalable implementation of GCNN called HydraGNN on a lattice of 256 atoms, which accounts well for the short-range interactions. Using this data, we predicted the formation energy for lattices of 864 atoms and 2048 atoms, which resulted in lower-than-expected accuracy due to the long-range interactions present in these larger lattices. We accounted for the long-range interactions by including a small amount of training data representative for those two larger sizes, whereupon the predictions of HydraGNN scaled linearly with the size of the lattice. Therefore, our strategy ensured scalability while reducing significantly the computational cost of training on larger lattice sizes.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Enhancing dimensionality prediction in hybrid metal halides via feature engineering and class-imbalance mitigation

We present a machine learning (ML) framework for predicting the structural dimensionality of hybrid metal halides (HMHs), including organic-inorganic perovskites, using a combination of chemically-informed feature engineering and advanced class-imbalance handling techniques. This study is motivated by the small and highly imbalanced nature of experimentally available HMH datasets, which limits the applicability and reliability of conventional ML approaches. The dataset, consisting of 494 HMH structures, is highly imbalanced across dimensionality classes (0D, 1D, 2D, 3D), posing significant challenges to predictive modeling. To mitigate this limitation, the dataset was augmented to 1336 samples using the synthetic minority oversampling technique, enabling improved learning of underrepresented dimensionality classes while preserving chemically meaningful feature relationships. We developed interaction-based descriptors designed to capture coupled steric and polarity effects relevant to dimensionality prediction, which are not readily captured by standard single-parameter or composition-only descriptors. These descriptors are integrated into a multi-stage workflow combining feature selection, ensemble stacking, and performance optimization. Our approach significantly improves F1-scores for underrepresented classes, achieving robust cross-validation performance across all dimensionalities. This work demonstrates a generalizable strategy for extracting reliable and interpretable structure–dimensionality relationships from limited experimental data, enabling pre-synthesis screening of organic cations and providing a practical blueprint for small-data ML in hybrid materials systems.

36 MATERIALS SCIENCE↗
DOE OSTI2023

The drivers and predictability of wildfire re-burns in the western United States (US)

Evidence is mounting that the effectiveness of using prescribed burns as a management tactic may be diminishing due to the higher incidence of wildfire re-burns. The development of predictive models of re-burns is thus essential to better understand their primary drivers so that forest management practices can be updated to account for these events. First, we assess the potential for human activity as a driver of re-burns by evaluating re-burn trends both within and outside of the wildland–urban interface (WUI) of the western US. Next, we investigate the predictability of re-burns through the application of both random forest and the explanatory machine learning non-negative matrix factorization using k-means clustering (NMFk) algorithms to predict re-burn occurrence over California based on a number of climate factors. Our findings indicate that while most states showed increasing trends within the WUI when trends were conducted over longer moving windows (e.g. 20 years), California was the only state where the rate of increase was consistently higher in the WUI, indicating a stronger potential for human activity as a driver in that location. Furthermore, we find model performance was found to be robust over most of California (Testing F1 scores = 0.688), although results were highly variable based on EPA level III Ecoregion (F1 scores = 0.0–0.778). Insights provided from this study will lead to a better understanding of climate and human activity drivers of re-burns and how these vary at broad spatial scales so that improvements in forest management practices can be tuned according to the level of change that is expected for a given region.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Integrating functional scoring and regulatory data to predict the effect of non-coding SNPs in a complex neurological disease

Abstract Most SNPs associated with complex diseases seem to lie in non-coding regions of the genome; however, their contribution to gene expression and disease phenotype remains poorly understood. Here, we established a workflow to provide assistance in prioritising the functional relevance of non-coding SNPs of candidate genes as susceptibility loci in polygenic neurological disorders. To illustrate the applicability of our workflow, we considered the multifactorial disorder migraine as a model to follow our step-by-step approach. We annotated the overlap of selected SNPs with regulatory elements and assessed their potential impact on gene expression based on publicly available prediction algorithms and functional genomics information. Some migraine risk loci have been hypothesised to reside in non-coding regions and to be implicated in the neurotransmission pathway. In this study, we used a set of 22 non-coding SNPs from neurotransmission and synaptic machinery-related genes previously suggested to be involved in migraine susceptibility based on our candidate gene association studies. After prioritising these SNPs, we focused on non-reported ones that demonstrated high regulatory potential: (1) VAMP2_rs1150 (3′ UTR) was predicted as a target of hsa-mir-5010-3p miRNA, possibly disrupting its own gene expression; (2) STX1A_rs6951030 (proximal enhancer) may affect the binding affinity of zinc-finger transcription factors (namely ZNF423) and disturb TBL2 gene expression; and (3) SNAP25_rs2327264 (distal enhancer) expected to be in a binding site of ONECUT2 transcription factor. This study demonstrated the applicability of our practical workflow to facilitate the prioritisation of potentially relevant non-coding SNPs and predict their functional impact in multifactorial neurological diseases.

Felício, Daniela↗
DOE OSTI2021

Artificial intelligence in the prediction of protein–ligand interactions: recent advances and future directions

Abstract New drug production, from target identification to marketing approval, takes over 12 years and can cost around $2.6 billion. Furthermore, the COVID-19 pandemic has unveiled the urgent need for more powerful computational methods for drug discovery. Here, we review the computational approaches to predicting protein–ligand interactions in the context of drug discovery, focusing on methods using artificial intelligence (AI). We begin with a brief introduction to proteins (targets), ligands (e.g. drugs) and their interactions for nonexperts. Next, we review databases that are commonly used in the domain of protein–ligand interactions. Finally, we survey and analyze the machine learning (ML) approaches implemented to predict protein–ligand binding sites, ligand-binding affinity and binding pose (conformation) including both classical ML algorithms and recent deep learning methods. After exploring the correlation between these three aspects of protein–ligand interaction, it has been proposed that they should be studied in unison. We anticipate that our review will aid exploration and development of more accurate ML-based prediction strategies for studying protein–ligand interactions.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Energy metric prediction for double insertion mutants via the RoseNet deep learning framework

Studying the structural and functional implications of protein mutations is an important task in computational biology and bioinformatics. We leverage our previously proposed RoseNet neural network architecture to predict energy metrics of proteins with double amino acid insertions or deletions (InDels). We train models on previously generated benchmark datasets containing the exhaustive double InDel mutations for three proteins, as well as an additional three proteins for which ∼145k random mutants, each with two InDels, have been generated. We expand on our previous work by evaluating three additional proteins and analyzing domain features that impact the prediction capabilities of RoseNet. These features include InDels into secondary structures and the solvent accessible surface area (SASA) scores of the residues. We uncover further evidence to support that RoseNet has a higher proficiency of generalizing to unseen residue combinations than unseen insertion positions. We also observe that RoseNet produces higher-quality predictions when inserting into a β-sheet over an α-helix. Additionally, when the insertions fall in an area of high SASA, RoseNet often displays better performance than inserting into areas of low SASA.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

EpitopeVec: linear epitope prediction using deep protein sequence embeddings

Motivation: B-cell epitopes (BCEs) play a pivotal role in the development of peptide vaccines, immuno-diagnostic reagents and antibody production, and thus in infectious disease prevention and diagnostics in general. Experimental methods used to determine BCEs are costly and time-consuming. Therefore, it is essential to develop computational methods for the rapid identification of BCEs. Although several computational methods have been developed for this task, generalizability is still a major concern, where cross-testing of the classifiers trained and tested on different datasets has revealed accuracies of 51–53%. Results: We describe a new method called EpitopeVec, which uses a combination of residue properties, modified antigenicity scales, and protein language model-based representations (protein vectors) as features of peptides for linear BCE predictions. Extensive benchmarking of EpitopeVec and other state-of-the-art methods for linear BCE prediction on several large and small datasets, as well as cross-testing, demonstrated an improvement in the performance of EpitopeVec over other methods in terms of accuracy and area under the curve. As the predictive performance depended on the species origin of the respective antigens (viral, bacterial and eukaryotic), we also trained our method on a large viral dataset to create a dedicated linear viral BCE predictor with improved cross-testing performance.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

The impact of curation errors in the PDBBind Database on machine learning predictions of protein–protein binding affinity

The PDBBind database has been widely utilized for the computational prediction of protein–protein binding affinities. While the accuracy of the PDBBind-curated equilibrium dissociation constants (K D ) has been reported for the protein–ligand subset of the PDBBind database, the curation accuracy has not been reported for the protein–protein subset. Here, we present a detailed manual analysis for the subset of PDBBind records with PubMed Central Open Access primary publications and find that ~19% of these records had K D values that were not supported by their primary publications. The impact of these putative curation errors on the machine learning-based prediction of K D from experimental protein–protein 3D structures was evaluated and correcting the curation errors improved the Pearson correlation coefficient between measured and random forest-predicted log 10 (K D ) values by ~8 percentage points. This finding underscores the importance of dataset accuracy for computational modelling and highlights the need for more stringent curation processes when extracting information from the scientific literature.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Integrating crop growth models with remote sensing for predicting biomass yield of sorghum

Abstract Plant phenotypes are often descriptive, rather than predictive of crop performance. As a result, extensive testing is required in plant breeding programmes to develop varieties aimed at performance in the target environments. Crop models can improve this testing regime by providing a predictive framework to (i) augment field phenotyping data and derive hard-to-measure phenotypes and (ii) estimate performance across geographical regions using historical weather data. The goal of this study was to parameterize the Agricultural Production Systems sIMulator (APSIM) crop growth models with remote-sensing and ground-reference data to predict variation in phenology and yield-related traits in 18 commercial grain and biomass sorghum hybrids. Genotype parameters for each hybrid were estimated using remote-sensing measurements combined with manual phenotyping in West Lafayette, IN, in 2018. The models were validated in hybrid performance trials in two additional seasons at that site and against yield trials conducted in Bushland, TX, between 2001 and 2018. These trials demonstrated that (i) maximum plant height, final dry biomass and radiation use efficiency (RUE) of photoperiod-sensitive and -insensitive forage sorghum hybrids tended to be higher than observed in grain sorghum, (ii) photoperiod-sensitive sorghum hybrids exhibited greater biomass production in longer growing environments and (iii) the parameterized and validated models perform well in above-ground biomass simulations across years and locations. Crop growth models that integrate remote-sensing data offer an efficient approach to parameterize larger plant breeding populations.

Yang, Kai-Wei↗
DOE OSTI2020

The predicted properties of helium-enriched globular cluster progenitors at high redshift

Globular cluster progenitors may have been detected by Hubble Space Telescope, and are predicted to be observable with James Webb Space Telescope (JWST) and ground-based extremely large telescopes with adaptive optics. This has the potential to elucidate the issue of globular cluster formation and the origins of significantly helium-enriched subpopulations, a problem in Galactic astronomy with no satisfactory theoretical solution. Given this context, we use model stellar tracks and isochrones to investigate the predicted observational properties of helium-enriched stellar populations in globular cluster progenitors. We find that, relative to helium-normal populations, helium-enriched (ΔY = +0.12) stellar populations similar to those inferred in the most massive globular clusters, are expected, modulo some rapid fluctuations in the first ~30 Myr, to be brighter and redder in the rest frame. At fixed age, stellar mass, and metallicity, a helium-enriched population is predicted to converge to being ~0.40 mag brighter at $\lambda \approx 2.0\, {\mu \rm m}$, and to be 0.30-mag redder in the JWST–NIRCam colour (F070W - F200W), and to actually be fainter for $\lambda \lesssim 0.50 \, {\mu \rm m}$. Separately, we find that the time-integrated shift in ionizing radiation is a negligible $\sim \!5{{\ \rm per\ cent}}$, though we show that the Lyman-α escape fraction could end up higher for helium-enriched stars.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Predicting bulge to total luminosity ratio of galaxies using deep learning

ABSTRACT We present a deep learning model to predict the r-band bulge-to-total luminosity ratio (B/T) of nearby galaxies using their multiband JPEG images alone. Our Convolutional Neural Network (CNN) based regression model is trained on a large sample of galaxies with reliable decomposition into the bulge and disc components. The existing approaches to estimate the B/T ratio use galaxy light-profile modelling to find the best fit. This method is computationally expensive, prohibitively so for large samples of galaxies, and requires a significant amount of human intervention. Machine learning models have the potential to overcome these shortcomings. In our CNN model, for a test set of 20 000 galaxies, 85.7 per cent of the predicted B/T values have absolute error (AE) less than 0.1. We see further improvement to 87.5 per cent if, while testing, we only consider brighter galaxies (with r-band apparent magnitude <17) with no bright neighbours. Our model estimates the B/T ratio for the 20 000 test galaxies in less than a minute. This is a significant improvement in inference time from the conventional fitting pipelines, which manage around 2–3 estimates per minute. Thus, the proposed machine learning approach could potentially save a tremendous amount of time, effort, and computational resources while predicting B/T reliably, particularly in the era of next-generation sky surveys such as the Legacy Survey of Space and Time (LSST) and the Euclid sky survey which will produce extremely large samples of galaxies.

Grover, Harsh (ORCID:0000000321338142)↗
DOE OSTI2022

How well do local relations predict gas-phase metallicity gradients? Results from SDSS-IV MaNGA

ABSTRACT Gas-phase metallicity gradients in galaxies provide important clues to those galaxies’ formation histories. Using SDSS-IV MaNGA data, we previously demonstrated that gas metallicity gradients vary systematically and significantly across the galaxy mass–size plane: at stellar masses beyond approximately $10^{10}\, \mathrm{M_\odot }$, more extended galaxies display steeper gradients (in units of dex/Re) at a given stellar mass. Here, we set out to develop a physical interpretation of these findings by examining the ability of local ∼kpc-scale relations to predict the gradient behaviour along the mass–size plane. We find that local stellar mass surface density, when combined with total stellar mass, is sufficient to reproduce the overall mass–size trend in a qualitative sense. We further find that we can improve the predictions by correcting for residual trends relating to the recent star formation histories of star-forming regions. However, we find as well that the most extended galaxies display steeper average gradients than predicted, even after correcting for residual metallicity trends with other local parameters. From these results, we argue that gas-phase metallicity gradients can largely be understood in terms of known local relations, but we also discuss some possible physical causes of discrepant gradients.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Predicting the Galactic population of free-floating planets from realistic initial conditions

ABSTRACT We present the first prediction for the mass distribution function of Galactic free-floating planets (FFPs) that aims to accurately include the relative contributions of multiple formation pathways and stellar populations. We derive our predicted distribution from dedicated simulations of planet birth, growth, migration, and ejection around circumbinary systems and extend these results to also include the contributions from single and wide binary systems. Our resulting FFP mass distribution shows several distinct features, including a strong peak at ${\sim} 8\, {\rm M}_{\oplus }$ arising from the transition between pebble and gas accretion regimes and a trough at ${\sim} 1\, {\rm M}_{\oplus }$ due to the shift in the dominant ejection process from planet–planet scattering to ejection through interactions with stars in circumbinary systems. We find that interactions with the central binary in close circumbinary systems are likely the dominant progenitor for FFPs more massive than Earth, leading to a steep power-law dependence in mass that agrees well with existing observations. In contrast, we find planet–planet scattering events in single and wide binary systems likely produce the majority of planets at Mars mass and below, resulting in a shallower power-law dependence. Our results suggest that existing extrapolations into the sub-terrestrial mass range may significantly overestimate the true FFP abundance. The features we predict in the mass distribution of FFPs will be detectable by upcoming space-based microlensing surveys and, if observed, will provide key insight into the origins of FFPs and the environments in which they form.

Coleman, Gavin A. L. (ORCID:0000000151118963)↗
DOE OSTI2021

Predictive Models of Genetic Redundancy in Arabidopsis thaliana

Abstract Genetic redundancy refers to a situation where an individual with a loss-of-function mutation in one gene (single mutant) does not show an apparent phenotype until one or more paralogs are also knocked out (double/higher-order mutant). Previous studies have identified some characteristics common among redundant gene pairs, but a predictive model of genetic redundancy incorporating a wide variety of features derived from accumulating omics and mutant phenotype data is yet to be established. In addition, the relative importance of these features for genetic redundancy remains largely unclear. Here, we establish machine learning models for predicting whether a gene pair is likely redundant or not in the model plant Arabidopsis thaliana based on six feature categories: functional annotations, evolutionary conservation including duplication patterns and mechanisms, epigenetic marks, protein properties including posttranslational modifications, gene expression, and gene network properties. The definition of redundancy, data transformations, feature subsets, and machine learning algorithms used significantly affected model performance based on holdout, testing phenotype data. Among the most important features in predicting gene pairs as redundant were having a paralog(s) from recent duplication events, annotation as a transcription factor, downregulation during stress conditions, and having similar expression patterns under stress conditions. We also explored the potential reasons underlying mispredictions and limitations of our studies. This genetic redundancy model sheds light on characteristics that may contribute to long-term maintenance of paralogs, and will ultimately allow for more targeted generation of functionally informative double mutants, advancing functional genomic studies.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

DeepNeo: a webserver for predicting immunogenic neoantigens

Abstract Non-self epitopes, whether originated from foreign substances or somatic mutations, trigger immune responses when presented by major histocompatibility complex (MHC) molecules and recognized by T cells. Identification of immunogenically active neoepitopes has significant implications in cancer and virus medicine. However, current methods are mostly limited to predicting physical binding of mutant peptides and MHCs. We previously developed a deep-learning based model, DeepNeo, to identify immunogenic neoepitopes by capturing the structural properties of peptide-MHC pairs with T cell reactivity. Here, we upgraded our DeepNeo model with up-to-date training data. The upgraded model (DeepNeo-v2) was improved in evaluation metrics and showed prediction score distribution that better fits known neoantigen behavior. The immunogenic neoantigen prediction can be conducted at https://deepneo.net.

Kim, Jeong Yeon↗
DOE OSTI2024

Predicting oncology drug-induced cardiotoxicity with donor-specific iPSC-CMs—a proof-of-concept study with doxorubicin

Abstract Many oncology drugs have been found to induce cardiotoxicity in a subset of patients, which significantly limits their clinical use and impedes the benefit of lifesaving anticancer treatments. Human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) carry donor-specific genetic information and have been proposed for exploring the interindividual difference in oncology drug-induced cardiotoxicity. Herein, we evaluated the inter- and intraindividual variability of iPSC-CM-related assays and presented a proof of concept to prospectively predict doxorubicin (DOX)-induced cardiotoxicity (DIC) using donor-specific iPSC-CMs. Our findings demonstrated that donor-specific iPSC-CMs exhibited greater line-to-line variability than the intraindividual variability in impedance cytotoxicity and transcriptome assays. The variable and dose-dependent cytotoxic responses of iPSC-CMs resembled those observed in clinical practice and largely replicated the reported mechanisms. By categorizing iPSC-CMs into resistant and sensitive cell lines based on their time- and concentration-related phenotypic responses to DOX, we found that the sensitivity of donor-specific iPSC-CMs to DOX may predict in vivo DIC risk. Furthermore, we identified a differentially expressed gene, DND microRNA-mediated repression inhibitor 1 (DND1), between the DOX-resistant and DOX-sensitive iPSC-CMs. Our results support the utilization of donor-specific iPSC-CMs in assessing interindividual differences in DIC. Further studies will encompass a large panel of donor-specific iPSC-CMs to identify potential novel molecular and genetic biomarkers for predicting DOX and other oncology drug-induced cardiotoxicity.

Toxicology↗