Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Curation”

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 703 records · Page 39
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Database of virus genomes from ultra-deep sequencing of wastewater

Researchers at University of Missouri have conducted ultra-deep RNA sequencing of viral concentrates from wastewater (1 billion Illumina reads per sample). The resulting dataset spans 321 samples collected weekly from 11 cities between 2023-2025. As part of a tri-lab collaboration, scientists at LLNL and LANL cleaned, assembled, and annotated this metagenomic data, identifying nearly 200,000 viral genomes. Careful data curation resulted in a database containing 21,015 high-quality, near-complete viral genomes from wastewater. This database contains viruses predicted to infect a range of hosts including bacteria (most common viruses), plants (most abundant viruses), and vertebrates (rarest viruses). There are also numerous novel viruses that could not be well identified and whose host(s) are unknown. Just 7% of all genomes in the wastewater virus database had genus-level matches in the public NCBI database, and 17% matched to a recently created metagenomic virus database at that level (metaVR). The database will provide baseline information about viruses in wastewater that may be used to additional identify novel viruses during ongoing monitoring

Allen, Jonathan [Lawrence Livermore National Labor↗
DOE OSTI2020

Detection of Head and Neck Cancer Based on Longitudinal Changes in Serum Protein Abundance

Head and Neck Squamous Cell Carcinoma (HNSCC) is significantly represented in the military population with substantial impact on military operations. Two-thirds of new HNSCC cases are first diagnosed at advanced stage III or IV with lymph node metastases. Thus, there is an urgent need of biomarkers for early detection and risk stratification. Archived serum samples from the Department of Defense Serum Repository provide a unique resource for longitudinal studies of HNSCC in the military and the identification of early detection biomarkers. Methods: We identified 175 patients diagnosed with HNSCC with 175 matched healthy controls and retrieved a total of 978 serum samples drawn at the time of diagnosis, 2 and 4 years prior to diagnosis, and 2 years after diagnosis. Following immunoaffinity depletion, serum samples were analyzed by targeted proteomics assays for multiplexed quantification of a panel of 146 candidate protein biomarkers from the curated literature. Results: We derived a 13-protein signature that distinguishes case versus controls based on longitudinal changes in serum protein concentration. The abundances of each of the 13 proteins remain constant over time in control subjects. The area under the curve for the derived random forest classifier was 0.90. Conclusions: This 13-protein classifier is highly promising for detection of HNSCC prior to overt symptoms. Impact: Use of longitudinal samples has significant potential to identify biomarkers for detection and risk stratification.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

LungMAP Portal Ecosystem: Systems-level Exploration of the Lung

An improved understanding of the human lung necessitates advanced systems models informed by an ever-increasing repertoire of molecular omics, cellular, imaging, and pathological datasets. To centralize and standardize information across broad lung research efforts we expanded the LungMAP.net website into a new gateway portal. This portal connects a broad spectrum of research networks, bulk and single-cell multi-omics data and a diverse collection of image data that span mammalian lung development, and disease. The data are standardized across species and technologies using harmonized data and metadata models that leverage recent advances including those from the Human Cell Atlas, diverse ontologies, and the LungMAP CellCards initiative. To cultivate future discoveries, we have aggregated a diverse collection of single-cell atlases for multiple species (human, rhesus, mouse), to enable consistent queries across technologies, cohorts, age, disease, and drug treatment. These atlases are provided as independent and integrated queryable datasets, with an emphasis on dynamic visualization, figure generation, re-analysis, cell-type curation, and automated reference-based classification of user-provided single-cell genomics datasets (Azimuth). Finally, as this resource grows, we intend to increase the breadth of available interactive interfaces, supported data types, data portals and datasets from LungMAP and external research efforts.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Exascale workflow applications and middleware: An ExaWorks retrospective

Exascale computers offer transformative capabilities to combine data-driven and learning-based approaches with traditional simulation applications to accelerate scientific discovery and insight. However, these software combinations and integrations are difficult to achieve due to the challenges of coordinating and deploying heterogeneous software components on diverse and massive platforms. Here, we present the ExaWorks project, which addresses many of these challenges. We developed a workflow Software Development Toolkit (SDK), a curated collection of workflow technologies that can be composed and interoperated through a common interface, engineered following current best practices, and specifically designed to work on HPC platforms. ExaWorks also developed PSI/J, a job management abstraction API, to simplify the construction of portable software components and applications that can be used over various HPC schedulers. The PSI/J API is a minimal interface for submitting and monitoring jobs and their execution state across multiple and commonly used HPC schedulers. We also describe several leading and innovative workflow examples of ExaWorks tools used on DOE leadership platforms. Furthermore, we discuss how our project is working with the workflow community, large computing facilities, and HPC platform vendors to address the requirements of workflows sustainably at the exascale.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Developing a SARS-CoV-2 main protease binding prediction random forest model for drug repurposing for COVID-19 treatment

The coronavirus disease 2019 (COVID-19) global pandemic resulted in millions of people becoming infected with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus and close to seven million deaths worldwide. It is essential to further explore and design effective COVID-19 treatment drugs that target the main protease of SARS-CoV-2, a major target for COVID-19 drugs. In this study, machine learning was applied for predicting the SARS-CoV-2 main protease binding of Food and Drug Administration (FDA)-approved drugs to assist in the identification of potential repurposing candidates for COVID-19 treatment. Ligands bound to the SARS-CoV-2 main protease in the Protein Data Bank and compounds experimentally tested in SARS-CoV-2 main protease binding assays in the literature were curated. These chemicals were divided into training (516 chemicals) and testing (360 chemicals) data sets. To identify SARS-CoV-2 main protease binders as potential candidates for repurposing to treat COVID-19, 1188 FDA-approved drugs from the Liver Toxicity Knowledge Base were obtained. A random forest algorithm was used for constructing predictive models based on molecular descriptors calculated using Mold2 software. Model performance was evaluated using 100 iterations of fivefold cross-validations which resulted in 78.8% balanced accuracy. The random forest model that was constructed from the whole training dataset was used to predict SARS-CoV-2 main protease binding on the testing set and the FDA-approved drugs. Model applicability domain and prediction confidence on drugs predicted as the main protease binders discovered 10 FDA-approved drugs as potential candidates for repurposing to treat COVID-19. Our results demonstrate that machine learning is an efficient method for drug repurposing and, thus, may accelerate drug development targeting SARS-CoV-2.

Research & Experimental Medicine↗
DOE OSTI2023

Systematic review of type 1 diabetes biomarkers reveals regulation in circulating proteins related to complement, lipid metabolism, and immune response

Type 1 diabetes (T1D) results from an autoimmune attack of the pancreatic β cells that progresses to dysglycemia and symptomatic hyperglycemia. Current biomarkers to track this evolution are limited, with development of islet autoantibodies marking the onset of autoimmunity and metabolic tests used to detect dysglycemia. Therefore, additional biomarkers are needed to better track disease initiation and progression. Multiple clinical studies have used proteomics to identify biomarker candidates. However, most of the studies were limited to the initial candidate identification, which needs to be further validated and have assays developed for clinical use. Here we curate these studies to help prioritize biomarker candidates for validation studies and to obtain a broader view of processes regulated during disease development. This systematic review was registered with Open Science Framework (https://doi.org/10.17605/OSF.IO/N8TSA). Using PRISMA guidelines, we conducted a systematic search of proteomics studies of T1D in the PubMed to identify putative protein biomarkers of the disease. Studies that performed mass spectrometry-based untargeted/targeted proteomic analysis of human serum/plasma of control, pre-seroconversion, post-seroconversion, and/or T1D-diagnosed subjects were included. For unbiased screening, 3 reviewers screened all the articles independently using the pre-determined criteria. A total of 13 studies met our inclusion criteria, resulting in the identification of 266 unique proteins, with 31 (11.6%) being identified across 3 or more studies. The circulating protein biomarkers were found to be enriched in complement, lipid metabolism, and immune response pathways, all of which are found to be dysregulated in different phases of T1D development. We found 2 subsets: 17 proteins (C3, C1R, C8G, C4B, IBP2, IBP3, ITIH1, ITIH2, BTD, APOE, TETN, C1S, C6A3, SAA4, ALS, SEPP1 and PI16) and 3 proteins (C3, CLUS and C4A) have consistent regulation in at least 2 independent studies at post-seroconversion and post-diagnosis compared to controls, respectively, making them strong candidates for clinical assay development. Biomarkers analyzed in this systematic review highlight alterations in specific biological processes in T1D, including complement, lipid metabolism, and immune response pathways, and may have potential for further use in the clinic as prognostic or diagnostic assays.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

DISTEMA: distance map-based estimation of single protein model accuracy with attentive 2D convolutional neural network

Abstract Background Estimation of the accuracy (quality) of protein structural models is important for both prediction and use of protein structural models. Deep learning methods have been used to integrate protein structure features to predict the quality of protein models. Inter-residue distances are key information for predicting protein’s tertiary structures and therefore have good potentials to predict the quality of protein structural models. However, few methods have been developed to fully take advantage of predicted inter-residue distance maps to estimate the accuracy of a single protein structural model. Result We developed an attentive 2D convolutional neural network (CNN) with channel-wise attention to take only a raw difference map between the inter-residue distance map calculated from a single protein model and the distance map predicted from the protein sequence as input to predict the quality of the model. The network comprises multiple convolutional layers, batch normalization layers, dense layers, and Squeeze-and-Excitation blocks with attention to automatically extract features relevant to protein model quality from the raw input without using any expert-curated features. We evaluated DISTEMA’s capability of selecting the best models for CASP13 targets in terms of ranking loss of GDT-TS score. The ranking loss of DISTEMA is 0.079, lower than several state-of-the-art single-model quality assessment methods. Conclusion This work demonstrates that using raw inter-residue distance information with deep learning can predict the quality of protein structural models reasonably well. DISTEMA is freely at https://github.com/jianlin-cheng/DISTEMA

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2019

tRNA functional signatures classify plastids as late-branching cyanobacteria

Eukaryotes acquired the trait of oxygenic photosynthesis through endosymbiosis of the cyanobacterial progenitor of plastid organelles. Despite recent advances in the phylogenomics of Cyanobacteria, the phylogenetic root of plastids remains controversial. Although a single origin of plastids by endosymbiosis is broadly supported, recent phylogenomic studies are contradictory on whether plastids branch early or late within Cyanobacteria. One underlying cause may be poor fit of evolutionary models to complex phylogenomic data. Using Posterior Predictive Analysis, we show that recently applied evolutionary models poorly fit three phylogenomic datasets curated from cyanobacteria and plastid genomes because of heterogeneities in both substitution processes across sites and of compositions across lineages. To circumvent these sources of bias, we developed CYANO-MLP, a machine learning algorithm that consistently and accurately phylogenetically classifies (“phyloclassifies”) cyanobacterial genomes to their clade of origin based on bioinformatically predicted function-informative features in tRNA gene complements. Classification of cyanobacterial genomes with CYANO-MLP is accurate and robust to deletion of clades, unbalanced sampling, and compositional heterogeneity in input tRNA data. CYANO-MLP consistently classifies plastid genomes into a late-branching cyanobacterial sub-clade containing single-cell, starch-producing, nitrogen-fixing ecotypes, consistent with metabolic and gene transfer data. Phylogenomic data of cyanobacteria and plastids exhibit both site-process heterogeneities and compositional heterogeneities across lineages. These aspects of the data require careful modeling to avoid bias in phylogenomic estimation. Furthermore, we show that amino acid recoding strategies may be insufficient to mitigate bias from compositional heterogeneities. However, the combination of our novel tRNA-specific strategy with machine learning in CYANO-MLP appears robust to these sources of bias with high accuracy in phyloclassification of cyanobacterial genomes. CYANO-MLP consistently classifies plastids as late-branching Cyanobacteria, consistent with independent evidence from signature-based approaches and some previous phylogenetic studies.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Identifying Genomic Islands with Deep Neural Networks

Background Horizontal gene transfer is the main source of adaptability for bacteria, through which genes are obtained from different sources including bacteria, archaea, viruses, and eukaryotes. This process promotes the rapid spread of genetic information across lineages, typically in the form of clusters of genes referred to as genomic islands (GIs). Different types of GIs exist, and are often classified by the content of their cargo genes or their means of integration and mobility. While various computational methods have been devised to detect different types of GIs, no single method is capable of detecting all types. Results We propose a method, which we call Shutter Island, that uses a deep learning model (Inception V3, widely used in computer vision) to detect genomic islands. The intrinsic value of deep learning methods lies in their ability to generalize. Via a technique called transfer learning, the model is pre-trained on a large generic dataset and then re-trained on images that we generate to represent genomic fragments. We demonstrate that this image-based approach generalizes better than the existing tools. Conclusions We used a deep neural network and an image-based approach to detect the most out of the correct GI predictions made by other tools, in addition to making novel GI predictions. The fact that the deep neural network was re-trained on only a limited number of GI datasets and then successfully generalized indicates that this approach could be applied to other problems in the field where data is still lacking or hard to curate.

Computer Vision↗
DOE OSTI2024

An evaluation of GPT models for phenotype concept recognition

Clinical deep phenotyping and phenotype annotation play a critical role in both the diagnosis of patients with rare disorders as well as in building computationally-tractable knowledge in the rare disorders field. These processes rely on using ontology concepts, often from the Human Phenotype Ontology, in conjunction with a phenotype concept recognition task (supported usually by machine learning methods) to curate patient profiles or existing scientific literature. With the significant shift in the use of large language models (LLMs) for most NLP tasks, we examine the performance of the latest Generative Pre-trained Transformer (GPT) models underpinning ChatGPT as a foundation for the tasks of clinical phenotyping and phenotype annotation. The experimental setup of the study included seven prompts of various levels of specificity, two GPT models (gpt-3.5-turbo and gpt-4.0) and two established gold standard corpora for phenotype recognition, one consisting of publication abstracts and the other clinical observations. The best run, using in-context learning, achieved 0.58 document-level F1 score on publication abstracts and 0.75 document-level F1 score on clinical observations, as well as a mention-level F1 score of 0.7, which surpasses the current best in class tool. Without in-context learning, however, performance is significantly below the existing approaches. Our experiments show that gpt-4.0 surpasses the state of the art performance if the task is constrained to a subset of the target ontology where there is prior knowledge of the terms that are expected to be matched. While the results are promising, the non-deterministic nature of the outcomes, the high cost and the lack of concordance between different runs using the same prompt and input make the use of these LLMs challenging for this particular task.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Chemical classification program synthesis using generative artificial intelligence

Accurately classifying chemical structures is essential for cheminformatics and bioinformatics, including tasks such as identifying bioactive compounds of interest, screening molecules for toxicity to humans, finding non-organic compounds with desirable material properties, or organizing large chemical libraries for drug discovery or environmental monitoring. However, manual classification is labor-intensive and difficult to scale to large chemical databases. Existing automated approaches either rely on manually constructed classification rules, or are deep learning methods that lack explainability. This work presents an approach that uses generative artificial intelligence to automatically write chemical classifier programs for classes in the Chemical Entities of Biological Interest (ChEBI) database. These programs can be used for efficient deterministic run-time classification of SMILES structures, with natural language explanations. The programs themselves constitute an explainable computable ontological model of chemical class nomenclature, which we call the ChEBI Chemical Class Program Ontology (C3PO). We validated our approach against the ChEBI database, and compared our results against deep learning models and a naive SMARTS pattern based classifier. C3PO outperforms the naive classifier, but does not reach the performance of state of the art deep learning methods. However, C3PO has a number of strengths that complement deep learning methods, including explainability and reduced data dependence. C3PO can be used alongside deep learning classifiers to provide an explanation of the classification, where both methods agree. The programs can be used as part of the ontology development process, and iteratively refined by expert human curators.

Artificial Intelligence↗
DOE OSTI2026

Benchmarking the performance of uncertainty quantification methods for neural network-based interatomic potentials

Machine-learned interatomic potentials (ML-IAPs) continue to gain popularity as accurate, computationally efficient replacements for traditional, physics-based interatomic potentials and expensive ab initio methods. Uncertainty quantification (UQ) of ML-IAPs is a growing area of research as UQ is critical in many applications of IAPs, such as developing curated datasets, active learning-based data augmentation, self-improving models, and estimating the uncertainty of molecular dynamics simulations. In this paper, we construct and benchmark a series of different neural network potentials (NNPs) with varying network architectures to determine the performance of these models with respect to both the mean and uncertainty calibration error. Each NNP method is specifically designed to predict either epistemic or aleatoric uncertainty with particular focus on the differences in behavior between the epistemic and aleatoric uncertainty estimates. We benchmark these methods using multiple datasets common in the ML-IAP literature. The results show that the aleatoric uncertainty from single-shot model architectures is a competitive alternative to ensemble-based epistemic uncertainty predictions in regions of sufficient data-density. However, in regions where the representative data is sparse, aleatoric uncertainty models tend to overpredict and epistemic methods tend to underpredict the actual model error. We conclude that the type of UQ is crucial when discussing performance of probabilistic model results as different methods have different performance characteristics depending on the regime in which they are evaluated. Therefore, the type of UQ method should be carefully evaluated against both the data characteristics and requirements for the intended application.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Minnesota peat viromes reveal terrestrial and aquatic niche partitioning for local and global viral populations

Abstract Background Peatlands are expected to experience sustained yet fluctuating higher temperatures due to climate change, leading to increased microbial activity and greenhouse gas emissions. Despite mounting evidence for viral contributions to these processes in peatlands underlain with permafrost, little is known about viruses in other peatlands. More generally, soil viral biogeography and its potential drivers are poorly understood at both local and global scales. Here, 87 metagenomes and five viral size-fraction metagenomes (viromes) from a boreal peatland in northern Minnesota (the SPRUCE whole-ecosystem warming experiment and surrounding bog) were analyzed for dsDNA viral community ecological patterns, and the recovered viral populations (vOTUs) were compared with our curated PIGEON database of 266,125 vOTUs from diverse ecosystems. Results Within the SPRUCE experiment, viral community composition was significantly correlated with peat depth, water content, and carbon chemistry, including CH 4 and CO 2 concentrations, but not with temperature during the first 2 years of warming treatments. Peat vOTUs with aquatic-like signatures (shared predicted protein content with marine and/or freshwater vOTUs) were significantly enriched in more waterlogged surface peat depths. Predicted host ranges for SPRUCE vOTUs were relatively narrow, generally within a single bacterial genus. Of the 4326 SPRUCE vOTUs, 164 were previously detected in other soils, mostly peatlands. None of the previously identified 202,371 marine and freshwater vOTUs in our PIGEON database were detected in SPRUCE peat, but 0.4% of 80,714 viral clusters (VCs, grouped by predicted protein content) were shared between soil and aquatic environments. On a per-sample basis, vOTU recovery was 32 times higher from viromes compared with total metagenomes. Conclusions Results suggest strong viral “species” boundaries between terrestrial and aquatic ecosystems and to some extent between peat and other soils, with differences less pronounced at higher taxonomic levels. The significant enrichment of aquatic-like vOTUs in more waterlogged peat suggests that viruses may also exhibit niche partitioning on more local scales. These patterns are presumably driven in part by host ecology, consistent with the predicted narrow host ranges. Although more samples and increased sequencing depth improved vOTU recovery from total metagenomes, the substantially higher per-sample vOTU recovery after viral particle enrichment highlights the utility of soil viromics.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

Exposing new taxonomic variation with inflammation — a murine model-specific genome database for gut microbiome researchers

The murine CBA/J mouse model widely supports immunology and enteric pathogen research. This model has illuminated Salmonella interactions with the gut microbiome since pathogen proliferation does not require disruptive pretreatment of the native microbiota, nor does it become systemic, thereby representing an analog to gastroenteritis disease progression in humans. Despite the value to broad research communities, microbiota in CBA/J mice are not represented in current murine microbiome genome catalogs. Here we present the first microbial and viral genomic catalog of the CBA/J murine gut microbiome. Using fecal microbial communities from untreated and Salmonella-infected, highly inflamed mice, we performed genomic reconstruction to determine the impacts on gut microbiome membership and functional potential. From high depth whole community sequencing (~ 42.4 Gbps/sample), we reconstructed 2281 bacterial and 4516 viral draft genomes. Salmonella challenge significantly altered gut membership in CBA/J mice, revealing 30 genera and 98 species that were conditionally rare and unsampled in non-inflamed mice. Additionally, inflamed communities were depleted in microbial genes that modulate host anti-inflammatory pathways and enriched in genes for respiratory energy generation. Our findings suggest decreases in butyrate concentrations during Salmonella infection corresponded to reductions in the relative abundance in members of the Alistipes. Strain-level comparison of CBA/J microbial genomes to prominent murine gut microbiome databases identified newly sampled lineages in this resource, while comparisons to human gut microbiomes extended the host relevance of dominant CBA/J inflammation-resistant strains. This CBA/J microbiome database provides the first genomic sampling of relevant, uncultivated microorganisms within the gut from this widely used laboratory model. Using this resource, we curated a functional, strain-resolved view on how Salmonella remodels intact murine gut communities, advancing pathobiome understanding beyond inferences from prior amplicon-based approaches. Salmonella-induced inflammation suppressed Alistipes and other dominant members, while rarer commensals like Lactobacillus and Enterococcus endure. The rare and novel species sampled across this inflammation gradient advance the utility of this microbiome resource to benefit the broad research needs of the CBA/J scientific community, and those using murine models for understanding the impact of inflammation on the gut microbiome more generally.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

BinaRena: a dedicated interactive platform for human-guided exploration and binning of metagenomes

Background: Exploring metagenomic contigs and “binning” them into metagenome-assembled genomes (MAGs) are essential for the delineation of functional and evolutionary guilds within microbial communities. Despite the advances in automated binning algorithms, their capabilities in recovering MAGs with accuracy and biological relevance are so far limited. Researchers often find that human involvement is necessary to achieve representative binning results. This manual process however is expertise demanding and labor intensive, and it deserves to be supported by software infrastructure. Results: We present BinaRena, a comprehensive and versatile graphic interface dedicated to aiding human operators to explore metagenome assemblies via customizable visualization and to associate contigs with bins. Contigs are rendered as an interactive scatter plot based on various data types, including sequence metrics, coverage profiles, taxonomic assignments, and functional annotations. Various contig-level operations are permitted, such as selection, masking, highlighting, focusing, and searching. Binning plans can be conveniently edited, inspected, and compared visually or using metrics including silhouette coefficient and adjusted Rand index. Completeness and contamination of user-selected contigs can be calculated in real time. In demonstration of BinaRena’s usability, we show that it facilitated biological pattern discovery, hypothesis generation, and bin refinement in a complex tropical peatland metagenome. It enabled isolation of pathogenic genomes within closely related populations from the gut microbiota of diarrheal human subjects. It significantly improved overall binning quality after curating results of automated binners using a simulated marine dataset. Conclusions: BinaRena is an installation-free, dependency-free, client-end web application that operates directly in any modern web browser, facilitating ease of deployment and accessibility for researchers of all skill levels. The program is hosted at https://github.com/qiyunlab/binarena, together with documentation, tutorials, example data, and a live demo. It effectively supports human researchers in intuitive interpretation and fine tuning of metagenomic data.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

FaceBase 3: analytical tools and FAIR resources for craniofacial and dental research

ABSTRACT The FaceBase Consortium was established by the National Institute of Dental and Craniofacial Research in 2009 as a ‘big data’ resource for the craniofacial research community. Over the past decade, researchers have deposited hundreds of annotated and curated datasets on both normal and disordered craniofacial development in FaceBase, all freely available to the research community on the FaceBase Hub website. The Hub has developed numerous visualization and analysis tools designed to promote integration of multidisciplinary data while remaining dedicated to the FAIR principles of data management (findability, accessibility, interoperability and reusability) and providing a faceted search infrastructure for locating desired data efficiently. Summaries of the datasets generated by the FaceBase projects from 2014 to 2019 are provided here. FaceBase 3 now welcomes contributions of data on craniofacial and dental development in humans, model organisms and cell lines. Collectively, the FaceBase Consortium, along with other NIH-supported data resources, provide a continuously growing, dynamic and current resource for the scientific community while improving data reproducibility and fulfilling data sharing requirements.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2019

Squamous Cell Carcinoma with Regional Metastasis to Axilla or Groin Lymph Nodes: a Multicenter Outcome Analysis

Background: Cutaneous squamous cell carcinoma (cSCC) of the trunk/extremities with nodal metastasis represents a rare but significant clinical challenge. Treatment patterns and outcomes are poorly described. Patients and Methods: Patients with cSCC who developed axilla/groin lymph node metastasis and underwent curative-intent surgery between 2005 and 2015 were identified at four Canadian academic centers. Demographics, tumor characteristics, treatment patterns, recurrence rates, and mortality were described. Overall survival (OS) and disease-free survival (DFS) were calculated using Kaplan–Meier analysis. Predictors of survival and any recurrence were explored using Cox regression and logistic regression models, respectively. Results: Of 43 patients, 70% were male (median age 74 years). Median follow-up was 38 months. Median time to nodal metastasis was 11.3 months. Thirty-one and 12 patients had nodal metastasis to the axilla and groin, respectively. A total of 72% and 7% received adjuvant and neoadjuvant radiation, respectively, while 5% received adjuvant chemotherapy. Following surgery, 26% patients developed nodal and/or distant disease recurrence. Crude mortality rate was 39.5%. Mean OS was 5.3 years [95% confidence interval (CI) 3.9–6.8 years], and 5-year OS was 55.1%. Mean DFS was 4.8 years (95% CI 3.3–6.2 years), and five-year DFS was 49.3%. Any recurrence was the only independent predictor of death [p = 0.036, odds ratio (OR) = 29.5], and extracapsular extension (p = 0.028, OR = 189) and age (p = 0.017, OR = 0.823) were independent predictors of recurrence. Conclusions: This represents the largest contemporary series to date of outcomes for patients with axilla/groin nodal metastases from cSCC. Despite aggressive treatment, outcomes remain modest, indicating the need for a continued multidisciplinary approach and integration of new systemic agents.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2024

HydraGNN_Predictive_GFM_2024 - Ensemble of predictive graph foundation models for ground state atomistic materials modeling

We provide the ensemble of fifteen pre-trained graph foundation models (GFMs) for atomistic materials modeling applications. Each one of the fifteen GFMs has been trained on five open-source datasets that (once aggregated) amount to over 154 million atomistic structures, which cover over two-thirds of the natural elements of the periodic table and that comprises a broad set of organic and inorganic compounds. This vast set of atomistic structures comprises ground state configurations that are dynamically stable (i.e., equilibrated structures with atomic forces approximately close to zero values) as well as dynamically unstable structures (i.e., non-equilibrium structures with non-negligible non-zero values of atomic forces). The ensemble of datasets aggregated does NOT include excited states. The datasets have been curated to remove atomistic structures with spectral norm of the force tensor above 100 eV/angstrom. Moreover, a linear term of the energy was computed for each dataset using a linear regression model that uses the chemical concentration of each natural element as regressor. The linear term predicted by the linear regression model has been subtracted from each original energy value to perform a re-alignment of the energy values across different electronic structures approximation theories performed to generate the diverse multi-source, multi-fidelity datasets. The folder "ADIOS_files" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "ADIOS_files" directory contains 6 sub-directories named as follows: - ANI1x-v3.bp - MPTrj-v3.bp - OC2020-20M-v3.bp - OC2020-v3.bp - OC2022-v3.bp - qm7x-v3.bp Each sub-directory contains the pre-processed datasets converted in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used to the development, training, and performance testing of the ensemble go predictive graph foundation models. Each GFM was developed using HydraGNN (https://github.com/ORNL/HydraGNN) as underlying graph neural network (GNN) architecture. The multi-task learning (MTL) capability of HydraGNN was used to simultaneously train the GFMs on labeled values for direct predictions of energy (a total system property of an atomistic structure that measures the chemical stability) and atomic forces (an atomic level property of an atomistic structure that measures the dynamical stability). The hyper parameters of the GFM have been tuned using scalable hyperparameter optimization (HPO) algorithms implemented in the software DeepHyper (https://github.com/deephyper/deephyper). The pre-training of each HPO trial was performed using distributed data parallelism (DDP) to scale the training across 128 compute nodes of the exascale OLCF supercomputer Frontier. Each HPO trial was trained only for 10 epochs and an early stopping was performed to avoid wasting significant computational resources on GNN architectures that were clearly underperforming. For each HPO trial, the 'omnistat' tool developed by (AMD Research - Advanced Micro Device) was used to measure the total energy consumption in kWh. The ensemble of GFMs was obtained by selecting the fifteen best performing HPO trials. Four models have been selected for their clear advantage in accuracy, and these are the GFMs with IDs 229, 156, 147, 260. Additional eleven models have been selected based on judicious balance between accuracy and energy consumption needed for training, and these are the GFMs with IDs 165, 78, 137, 1, 175, 171, 181, 67, 179, 167, 351. Each selected GFM of the ensemble was continued to cumulate a total of at most 30 epochs. In some cases, the total number of epochs actually performed was les than 30 due to two combined factors: (1) the size of the GFM (i.e., the number of model parameters to train) and (2) the total wall-clock time for which the computational resources could be allocated on OLCF-Frontier. The "Ensemble_of_models" directory contains 15 sub-directories named as follows: - gfm_0.229 - gfm_0.156 - gfm_0.147 - gfm_0.260 - gfm_0.165 - gfm_0.78 - gfm_0.137 - gfm_0.1 - gfm_0.175 - gfm_0.171 - gfm_0.181 - gfm_0.67 - gfm_0.179 - gfm_0.167 - gfm_0.351 Each one of these sub-directories refers to one of the fifteen HPO trials that have been selected to continue the pre-training with at most 30 epochs. With each sub-directory associated with a specific HPO trial, the following files can be found: - config.json: file for argument parsing to develop and train an HydraGNN architecture - gfm_0.ID_epoch_N.pk: file with model parameters for HPO ID trial after N epochs of training The ensemble of fifteen GFM architectures was used for (1) ensemble averaging to stabilize the predictions of energy and atomic forces after pre-training for post-processing analysis and (2) ensemble uncertainty quantification (UQ). The code used to develop, pre-train, and load the pre-trained models for post-processing analysis is available on the ORNL-GitHub at the following link: https://github.com/ORNL/HydraGNN/tree/Predictive_GFM_2024

36 MATERIALS SCIENCE↗