Results for “Library” — Page 35

DOE OSTI2024

Parallel Runtime Interface for Fortran (PRIF): A Multi-Image Solution for LLVM Flang

Fortran compilers that provide support for Fortran’s native parallel features often do so with a runtime library that depends on details of both the compiler implementation and the communication library, while others provide limited or no support at all. This paper introduces a new generalized interface that is both compiler- and runtime-library-agnostic, providing flexibility while fully supporting all of Fortran’s parallel features. The Parallel Runtime Interface for Fortran (PRIF) was developed to be portable across shared- and distributed-memory systems, with varying operating systems, toolchains and architectures. It achieves this by defining a set of Fortran procedures corresponding to each of the parallel features defined in the Fortran standard that may be invoked by a Fortran compiler and implemented by a runtime library. PRIF aims to be used as the solution for LLVM Flang to provide parallel Fortran support. This paper also briefly describes our PRIF prototype implementation: Caffeine.

Bonachea, Dan↗

DOE OSTI2024

A Full-Stack Exploration of Language-Based Parallelism in Fortran 2023

This poster explores native parallel features in Fortran 2023 through the lens of supporting applications with libraries, compilers, and parallel runtimes. The language revision informally named Fortran 2008 introduced parallelism in the form of Single Program Multiple Data (SPMD) execution with two broad feature sets: (1) loop-level parallelism via do concurrent and (2) a Partitioned Global Address Space (PGAS) comprised of distributed “coarray” data structures. Fortran’s native parallelism has demonstrated high performance [1] and reduced the burden of inserting what sometimes amounts to more directives than code. Several compilers support both feature sets, typically by translating do concurrent into serial do loops annotated by parallel directives and by translating SPMD/PGAS features into direct calls to a communication library. Our research focuses primarily on two questions: (1) can the compiler’s parallel runtime library be developed in the language being compiled (Fortran) and (2) can we define an interface to the runtime that liberates compilers from being hardwired to one runtime and vice versa. We are answering these questions by developing the Parallel Runtime Interface for Fortran (PRIF) [2] and the Co-Array Fortran Framework of Efficient Interfaces to Network Environments (Caffeine) [3]. Caffeine is initially targeting adoption by LLVM Flang, a new open-source Fortran compiler developed by a broad community in industry, academia, and government labs. We are also exploring the use of these features in Inference-Engine, a deep learning library designed to facilitate neural network training and inference for high-performance computing applications written in modern Fortran.

Rasmussen, Katherine↗

DOE OSTI2024

KBase Narrative - Porphyromonadaceae sp. W3.11 genome

Narratives for The phenotype and genotype of fermentative prokaryotes This is the Narrative for Porphyromonadaceae sp. W3.11. A complementary Narrative for Lachnospiraceae sp. C1.1 is available here. This is the Narrative for Lachnospiraceae sp. C1.1. A complementary Narrative for Porphyromonadaceae sp. W3.11 is available here. Background and Isolation This Narrative and its complementary Narrative contain assembly and annotation of two bacterial isolates that were isolated by our laboratory from the rumen of a Holstein heifer. All procedures with animals have been approved by University of California Davis’s Institutional Animal Care and Use Committee. Rumen contents were collected through a rumen fistula and strained through two layers of cheesecloth into a bottle. The bottle was sealed to exclude air and maintained at 39°C. Contents were brought to the laboratory and bubbled under O2-free CO2 within 15 min. At the laboratory, serial dilutions were made with anaerobic dilution solution for Lachnospiraceae sp. C1.1 and propionibacterium diluent for Porphyromonadaceae sp. W3.11 (table S2). Aliquots (0.1 ml) of each dilution were injected into anaerobic bottle plates (1) containing 9 ml of LH medium (table S2). After incubation at 37°C for 7 days, isolated colonies were picked. Lachnospiraceae sp. C1.1 was picked from a bottle inoculated with a 104 dilution of rumen contents, and Porphyromonadaceae sp. W3.11 was picked from a bottle inoculated with a 103 dilution. After initial isolation, these organisms were purified by growing on anaerobic roll tubes (2) and picking isolated colonies. We performed de novo sequencing of Lachnospiraceae sp. C1.1 and Porphyromonadaceae sp. W3.11. Aliquots of liquid culture (9 and 1.5 ml, respectively) were collected by syringe and centrifuged (21,000g for 10 min at 4°C). Cell pellets were submitted to Molecular Research LP for DNA extraction, library preparation, and sequencing. After resuspending pellets in 180 µl of ATL buffer (Qiagen), DNA was extracted using the MagAttract HMW DNA Kit (Qiagen). DNA was eluted in 100 µl of AE buffer (Qiagen) and then cleaned using the DNEasy PowerClean Pro Cleanup Kit (Qiagen). DNA was then sheared using the Covaris g-TUBE (Covaris). Sequencing libraries were prepared using the SMRTbell Express Template Prep Kit 2.0 (Pacific Biosciences) and 1500 ng of the sheared and purified DNA. The SMRTbell libraries were size-selected (>6 Kb) using a BluePippin instrument (Sage Science) and 0.75% agarose gel. Libraries were then sequenced using the PacBio Sequel II (Pacific Biosciences) platform and a 30-hour movie time. Narrative Summary In these Narratives, we filtered low-quality reads using Trimmomatic (v0.36), assembled filtered reads with SPAdes (v3.15.3), and then checked completeness and contamination of the assembled genomes with CheckM (v1.0.18). Statistics for sequencing and assembly are in table S3. Using the assembled contigs (genomes), we called genes and annotated them. Protein-coding genes were called using Prodigal (v2.6.3) (3) locally or using KBase via RASTtk (v1.073), with identical results. Genes were annotated with KO IDs using KAAS (4). They were further annotated with pfam and TIGRFAM IDs using KBase and the Annotate Domains in a Genome app. We classified putative genes for hydrogenases using HydDB. Genes for 16S ribosomal RNA (rRNA) were called using RASTtk (v1.073) in KBase. The contigs (genomes) were analyzed to determine whether they belonged to new species. Taxonomy was assigned using GTDB-Tk (v1.7.0) in KBase. The identity of 16S rRNA genes to other organisms was found using EzBioCloud (5). Values of digital DNA-DNA hybridization (dDDH) were found with Type (Strain) Genome Server (6). These analyses suggest that Lachnospiraceae sp. C1.1 and Porphyromonadaceae sp. W3.11 represent novel species or genera. GTDB-Tk assigned Lachnospiracae sp. C1.1 to family Lachnospiraceae and genus NK4A144, which contains no type strains. It assigned Porphyromonadaceae sp. W3.11 to Porphyromonadaceae and genus Porphyromonas_A. Values of 16S rRNA identity and dDDH with respect to type strains were low (table S4). Although more phenotypic data are needed, available evidence supports assignment of genomes to new species or genera. Related publication Hackmann TJ, Zhang B. The phenotype and genotype of fermentative prokaryotes. Sci Adv. 2023 Sep 29;9(39):eadg8687. doi: 10.1126/sciadv.adg8687. Epub 2023 Sep 27. PMID: 37756392; PMCID: PMC10530074.

Hackmann, Timothy↗

DOE OSTI2024

KBase Narrative - Lachnospiraceae sp. C1.1 genome

Narratives for The phenotype and genotype of fermentative prokaryotes This is the Narrative for Porphyromonadaceae sp. W3.11. A complementary Narrative for Lachnospiraceae sp. C1.1 is available here. This is the Narrative for Lachnospiraceae sp. C1.1. A complementary Narrative for Porphyromonadaceae sp. W3.11 is available here. Background and Isolation This Narrative and its complementary Narrative contain assembly and annotation of two bacterial isolates that were isolated by our laboratory from the rumen of a Holstein heifer. All procedures with animals have been approved by University of California Davis’s Institutional Animal Care and Use Committee. Rumen contents were collected through a rumen fistula and strained through two layers of cheesecloth into a bottle. The bottle was sealed to exclude air and maintained at 39°C. Contents were brought to the laboratory and bubbled under O2-free CO2 within 15 min. At the laboratory, serial dilutions were made with anaerobic dilution solution for Lachnospiraceae sp. C1.1 and propionibacterium diluent for Porphyromonadaceae sp. W3.11 (table S2). Aliquots (0.1 ml) of each dilution were injected into anaerobic bottle plates (1) containing 9 ml of LH medium (table S2). After incubation at 37°C for 7 days, isolated colonies were picked. Lachnospiraceae sp. C1.1 was picked from a bottle inoculated with a 104 dilution of rumen contents, and Porphyromonadaceae sp. W3.11 was picked from a bottle inoculated with a 103 dilution. After initial isolation, these organisms were purified by growing on anaerobic roll tubes (2) and picking isolated colonies. We performed de novo sequencing of Lachnospiraceae sp. C1.1 and Porphyromonadaceae sp. W3.11. Aliquots of liquid culture (9 and 1.5 ml, respectively) were collected by syringe and centrifuged (21,000g for 10 min at 4°C). Cell pellets were submitted to Molecular Research LP for DNA extraction, library preparation, and sequencing. After resuspending pellets in 180 µl of ATL buffer (Qiagen), DNA was extracted using the MagAttract HMW DNA Kit (Qiagen). DNA was eluted in 100 µl of AE buffer (Qiagen) and then cleaned using the DNEasy PowerClean Pro Cleanup Kit (Qiagen). DNA was then sheared using the Covaris g-TUBE (Covaris). Sequencing libraries were prepared using the SMRTbell Express Template Prep Kit 2.0 (Pacific Biosciences) and 1500 ng of the sheared and purified DNA. The SMRTbell libraries were size-selected (>6 Kb) using a BluePippin instrument (Sage Science) and 0.75% agarose gel. Libraries were then sequenced using the PacBio Sequel II (Pacific Biosciences) platform and a 30-hour movie time. Narrative Summary In these Narratives, we filtered low-quality reads using Trimmomatic (v0.36), assembled filtered reads with SPAdes (v3.15.3), and then checked completeness and contamination of the assembled genomes with CheckM (v1.0.18). Statistics for sequencing and assembly are in table S3. Using the assembled contigs (genomes), we called genes and annotated them. Protein-coding genes were called using Prodigal (v2.6.3) (3) locally or using KBase via RASTtk (v1.073), with identical results. Genes were annotated with KO IDs using KAAS (4). They were further annotated with pfam and TIGRFAM IDs using KBase and the Annotate Domains in a Genome app. We classified putative genes for hydrogenases using HydDB. Genes for 16S ribosomal RNA (rRNA) were called using RASTtk (v1.073) in KBase. The contigs (genomes) were analyzed to determine whether they belonged to new species. Taxonomy was assigned using GTDB-Tk (v1.7.0) in KBase. The identity of 16S rRNA genes to other organisms was found using EzBioCloud (5). Values of digital DNA-DNA hybridization (dDDH) were found with Type (Strain) Genome Server (6). These analyses suggest that Lachnospiraceae sp. C1.1 and Porphyromonadaceae sp. W3.11 represent novel species or genera. GTDB-Tk assigned Lachnospiracae sp. C1.1 to family Lachnospiraceae and genus NK4A144, which contains no type strains. It assigned Porphyromonadaceae sp. W3.11 to Porphyromonadaceae and genus Porphyromonas_A. Values of 16S rRNA identity and dDDH with respect to type strains were low (table S4). Although more phenotypic data are needed, available evidence supports assignment of genomes to new species or genera. Related publication Hackmann TJ, Zhang B. The phenotype and genotype of fermentative prokaryotes. Sci Adv. 2023 Sep 29;9(39):eadg8687. doi: 10.1126/sciadv.adg8687. Epub 2023 Sep 27. PMID: 37756392; PMCID: PMC10530074.

Hackmann, Timothy↗

DOE OSTI2023

gRNA-SeqRET: a universal tool for targeted and genome-scale gRNA design and sequence extraction for prokaryotes and eukaryotes

High-throughput genetic screening is frequently employed to rapidly associate gene with phenotype and establish sequence-function relationships. With the advent of CRISPR technology, and the ability to functionally interrogate previously genetically recalcitrant organisms, non-model organisms can be investigated using pooled guide RNA (gRNA) libraries and sequencing-based assays to quantitatively assess fitness of every targeted locus in parallel. To aid the construction of pooled gRNA assemblies, we have developed an in silico design workflow for gRNA selection using the gRNA Sequence Region Extraction Tool (gRNA-SeqRET). Built upon the previously developed CCTop, gRNA-SeqRET enables automated, scalable design of gRNA libraries that target user-specified regions or whole genomes of any prokaryote or eukaryote. Additionally, gRNA-SeqRET automates the bulk extraction of any regions of sequence relative to genes or other features, aiding in the design of homology arms for insertion or deletion constructs. We also assess in silico the application of a designed gRNA library to other closely related genomes and demonstrate that for very closely related organisms Average Nucleotide Identity (ANI) > 95% a large fraction of the library may be of relevance. The gRNA-SeqRET web application pipeline can be accessed at https://grna.jgi.doe.gov. The source code is comprised of freely available software tools and customized Python scripts, and is available at https://bitbucket.org/berkeleylab/grnadesigner/src/master/ under a modified BSD open-source license (https://bitbucket.org/berkeleylab/grnadesigner).

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2021

The Effect of the Flux Separability Approximation on Multigroup Neutron Transport

The angular dependence of flux-weighted multigroup cross sections is commonly neglected when generating multigroup libraries. The error of this flux separability approximation is typically not isolated from other error sources due to a lack of availability of library generation and corresponding solvers that cannot relax this approximation. These errors can now be isolated and quantified with the availability of a multigroup Monte Carlo transport and multigroup library-generation capability in the OpenMC Monte Carlo transport code. This work will discuss relevant details of the OpenMC implementation, provide an example case useful for detailing the type of errors one can expect from making the flux separability approximation, and end with more realistic problems which show the impact of the approximation and highlight how it can strongly arise from an energy-dependent resonance absorption effect. Since the angle-dependence is intrinsically linked to the energy group structure, these examples also show that relaxing the flux separability approximation with angle-dependent cross sections could be used to reduce either the fine-tuning required to set a multigroup energy structure for a specific reactor type or the number of energy groups required to obtain a desired level of accuracy for a given problem. This trade-off could increase the costs of generating multigroup cross sections, and has the potential to require more memory for storing the multigroup library during the transport calculations, but it can significantly reduce the computational time required since the runtime of a discrete ordinates or method of characteristics neutron transport solver scales roughly linearly with the number of groups.

42 ENGINEERING↗

DOE OSTI2021

Generation of the TSL for Zirconium Hydrides from Ab Initio Methods

Zirconium hydride (ZrHx) is a moderator material used in TRIGA and other reactors that may exist in multiple phases with varying stoichiometry, which include the δ phase and the ϵ phase. Current ENDF/B-VIII.0 ZrHx thermal scattering law (TSL) evaluations do not distinguish between phases. These sub-libraries were generated with the LEAPR module of NJOY using historic phonon spectra derived from a central force model and assume incoherent elastic scattering for both bound hydrogen and zirconium, which neglects the effects of crystal structures important for scattering from zirconium bound in ZrHx. In this work, the TSLs for hydrogen and zirconium bound in δ-ZrHx and ϵ-ZrH2 were generated from phonon spectra derived from modern ab initio lattice dynamics methods and ab initio molecular dynamics. Subsequently, TSLs for hydrogen and zirconium in ZrHx and ZrH2 were generated using the Full Law Analysis Scattering System Hub (FLASSH) code. The built-in generalized coherent elastic routine was used to generate the previously neglected elastic contribution from zirconium for this material. The present TSLs provide both a re-evaluation of the current ZrH sub-libraries and expansion of the set of TSLs available for the examination of neutrons in systems with zirconium hydride, permitting explicit treatment of δ and ϵ phases.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2023

Xarray Climate Data Analysis Tools

xCDAT is an extension of xarray for climate data analysis on structured grids. It serves as a modern successor to the Community Data Analysis Tools (CDAT) library. Xarray is an "open source project and Python package that introduces labels in the form of dimensions, coordinates, and attributes on top of raw NumPy-like arrays, which allows for more intuitive, more concise, and less error-prone user experience. Xarray includes a large and growing library of domain-agnostic functions for advanced analytics and visualization with these data structures" (source: https://xarray.dev/). The goal of xCDAT is to provide generalizable features and utilities for simple and robust analysis of climate data. xCDAT's design philosophy is focused on reducing the overhead required to accomplish certain tasks in xarray. Some key xCDAT features are inspired by or ported from the core CDAT library, while others leverage powerful libraries in the xarray ecosystem (e.g., xESMF and cf_xarray) to deliver robust APIs.

Vo, Tom↗

DOE OSTI2021

System and method of storing and analyzing information

A system and method of storing and analyzing information is disclosed. The system includes a compiler layer to convert user queries to data parallel executable code. The system further includes a library of multithreaded algorithms, processes, and data structures. The system also includes a multithreaded runtime library for implementing compiled code at runtime. The executable code is dynamically loaded on computing elements and contains calls to the library of multithreaded algorithms, processes, and data structures and the multithreaded runtime library.

Feo, John T.↗

DOE OSTI2022

Studying Performance Portability of LAMMPS across Diverse GPU-based Platforms

The molecular dynamics simulation software, LAMMPS, utilizes the Kokkos acceleration library to port computation to a diverse set of architectures including those based on GPU accelerators. In addition to Kokkos, LAMMPS contains a vast code base that leverages the CUDA application programming interface using library functions such as cuFFT, CUDA’s fast-fourier transform (FFT) library, and, more recently, also support for AMD’s Heterogeneous Interface for Portability (HIP) that is rapidly growing. While preparing LAMMPS tests for the AMD GPU-based test system precursors to Frontier, we investigated several strategies for accelerating LAMMPS on AMD GPUs, using the AMD Instinct MI100 and MI250X. In this work, we integrated the HIP FFT library, hipFFT, into the particle-particle particle-mesh (PPPM) long-range solver, which allowed the porting of PPPM calculations to the GPUs. Kokkos behavior on the MI100 and MI250X was also investigated through the package kokkos command of LAMMPS, targeting com- munication, memory usage, and particle grid decomposition. The Tersoff, Reax, Lennard-Jones (LJ), EAM, Granular, and PPPM potentials were investigated in this effort, and results from these experiments are provided. The selected potentials were run on Spock (AMD Instinct MI100), Crusher (AMD Instinct MI250X), AFW HPC11 (NVIDIA A100) and Summit (NVIDIA V100), for comparison. Operational roofline models were constructed and analyzed for the Tersoff, Reax, and Lennard-Jones potentials on Crusher and Summit.

Hagerty, Nick↗

DOE OSTI2022

Whole core analysis of OECD/NEA TVA-WB1 benchmark exercise 1 using continuous-energy Monte Carlo code MVP with JENDL-4.0u and ENDF/B-VII.0

The TVA WB1 benchmark problem was precisely modeled in accordance with the geometry and material specifications, and benchmark exercise 1 was solved with the MVP code and the nuclear data libraries based on JENDL-4.0u (JENDL-4.0 updated file) and ENDF/B-VII.0. The criticality, the control rod bank worth, the isothermal temperature coefficient (ITC), the integral rod worth of bank D, and differential soluble boron worth (DBW) were obtained. The criticality and the control rod bank worth were compared with the measured data and published calculation results with other codes, Serpent and KENO. The criticality and the control rod bank worth calculated by MVP and the libraries agree with the measured and the published calculation results within the criteria specified in the ANSI/ANS standard. From these results, the calculation model for the MVP and the libraries was verified. The other parameters such as ITC, the integral rod worth, and DBW were also compared between JENDL-4.0u and ENDF/B-VII.0. The comparison of the results shows JENDL-4.0u and ENDF/B-VII.0 give comparable results within the standard deviation, the difference between libraries is small. (author)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

DOE OSTI2024

autoGEMM: Pushing the Limits of Irregular Matrix Multiplication on Arm Architectures

This paper presents an open-source library that pushes the limits of performance portability for irregular General Matrix Multiplication (GEMM) on the widely-used Arm architectures. Our library, autoGEMM, is designed to support a wide range of Arm processors: from edge devices to HPC-grade CPUs. autoGEMM generates optimized kernels for various hardware configurations by auto-combining fragments of autogenerated micro-kernels that employ hand-written optimizations to maximize computational efficiency. We optimize the kernel pipeline by tuning the register reuse and the data load/store overlapping. In addition, we use a dynamic tiling scheme to generate balanced tile shapes. Finally, we position autoGEMM on top of the TVM framework where our dynamic tiling scheme prunes the search space for TVM to identify the optimal combination of parameters for code optimization. Evaluations on five different classes of Arm chips demonstrate the advantages of autoGEMM. For small matrices, autoGEMM achieves 98% of peak and up to 2.0x speedup over state-of-the-art libraries such as LIBXSMM and LibShalom. For irregular matrices (i.e. tall skinny and long rectangles), autoGEMM is 1.3-2.0x faster than widely-used libraries such as OpenBLAS and Eigen. autoGEMM is publicly available at: https://github.com/wudu98/autoGEMM.

Wu, Du↗

DOE OSTI2024

Thermo-Fluid Modeling Framework for Supercomputer Digital Twins: Part 1, Demonstration at Exascale

A thermo-fluid modeling framework is being developed for ExaDigiT---an open-source framework for developing comprehensive digital twins of liquid-cooled supercomputers. The work is being conducted in two parts, and discussion is divided into two companion papers. The work documented in this paper focuses on the development of a cooling system library in Dymola for the Frontier supercomputer at Oak Ridge National Laboratory. The second part, outlined in a companion paper, focuses on a templating structure called Auto-CSM for easily creating model-agnostic, physics-based thermo-fluid cooling system models for liquid-cooled supercomputers using a text-based schema. The cooling model is being developed using primarily the open-source Transient Simulation Framework of Reconfigurable Models (TRANSFORM) library. The library follows the templating architecture developed within the TRANSFORM library for modeling subsystems. A full-system validation was performed to validate a very simple model that is integrated with the system controls, and the results are presented herein.

Kumar, Vineet↗

DOE OSTI2022

Cited Reference Analysis in a National Laboratory to Assess Journal Usage

Librarians at a large, federally funded, multidisciplinary research organization sought to assess the value provided by the library’s collections. One method for capturing this value includes determining how well their library’s collections meet researcher needs, as expressed by the materials those researchers choose to cite in the production of new scholarly communications. A cited reference analysis was performed on the publications originating from the institution to provide the necessary data for the assessment, as well as to aid in the creation of a core list of journals which will inform future collection development efforts. Overall, this paper fills a gap in the library and information science literature by addressing the citation behaviors of researchers at a large multidisciplinary government research facility, a setting that has not been examined previously.

99 GENERAL AND MISCELLANEOUS↗

DOE OSTI2023

Navigating the Expansive Landscapes of Soft Materials: A User Guide for High-Throughput Workflows

Synthetic polymers are highly customizable with tailored structures and functionality, yet this versatility generates challenges in the design of advanced materials due to the size and complexity of the design space. Thus, exploration and optimization of polymer properties using combinatorial libraries has become increasingly common, which requires careful selection of synthetic strategies, characterization techniques, and rapid processing workflows to obtain fundamental principles from these large data sets. Herein, we provide guidelines for strategic design of macromolecule libraries and workflows to efficiently navigate these high-dimensional design spaces. We describe synthetic methods for multiple library sizes and structures as well as characterization methods to rapidly generate data sets, including tools that can be adapted from biological workflows. We further highlight relevant insights from statistics and machine learning to aid in data featurization, representation, and analysis. This Perspective acts as a “user guide” for researchers interested in leveraging high-throughput screening toward the design of multifunctional polymers and predictive modeling of structure–property relationships in soft materials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2023

Coupling High-Throughput and Targeted Screening for Identification of Nonobvious Metabolic Engineering Targets

Identification of metabolic engineering targets is a fundamental challenge in strain development programs. While high-throughput (HTP) genetic engineering methodologies capable of generating vast diversity are being developed at a rapid rate, a majority of industrially interesting molecules cannot be screened at sufficient throughput to leverage these techniques. We propose a workflow that couples HTP screening of common precursors (e.g., amino acids) that can be screened either directly or by artificial biosensors, with low-throughput targeted validation of the molecule of interest to uncover nonintuitive beneficial metabolic engineering targets and combinations hereof. Using this workflow, we identified several nonobvious novel targets for improving p-coumaric acid (p-CA) and L -DOPA production from two large 4k gRNA libraries each deregulating 1000 metabolic genes in the yeast Saccharomyces cerevisiae. We initially screened yeast cells transformed with gRNA library plasmids for individual regulatory targets improving the production of l-tyrosine-derived betaxanthins, identifying 30 targets that increased intracellular betaxanthin content 3.5–5.7 fold. Hereafter, we screened the targets individually in a high-producing p-CA strain, narrowing down the targets to six that increased the secreted titer by up to 15%. To investigate whether any of the six targets could be additively combined to improve p-CA production further, we created a gRNA multiplexing library and subjected it to our proposed coupled workflow. The combination of regulating PYC1 and NTH2 simultaneously resulted in the highest (threefold) improvement of the betaxanthin content, and an additive trend was also observed in the p-CA strain. Lastly, we tested the initial 30 targets in a L -DOPA producing strain, identifying 10 targets that increased the secreted titer by up to 89%, further validating our screening by proxy workflow. This coupled approach is useful for strain development in the absence of direct HTP screening assays for products of interest.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2022

Automatic Generation of High-Performance Convolution Kernels on ARM CPUs for Deep Learning

In this work, we present FastConv, a template-based code auto-generation open source library that can automatically generate high-performance deep learning convolution kernels of arbitrary matrices/tensors shapes. FastConv is based on the Winograd algorithm, which is reportedly the highest performing algorithm for the time-consuming convolution layers of convolutional neural networks. ARM CPUs cover a wide range designs and specifications, from embedded devices to HPC-grade CPUs. The leads to the dilemma of how to consistently optimize Winograd-based convolution solvers for convolution layers of different shapes. FastConv addresses this problem by using templates to auto-generate multiple shapes of tuned kernels variants suitable for skinny tall matrices. As a performance portable library, FastConv transparently searches for the best combination of kernel shapes, cache tiles, scheduling of loop orders, packing strategies, access patterns, and online/offline computations. Auto-tuning is used to search the parameter configuration space for the best performance for a given target architecture and problem size. The experiments with layer-wise evaluation on the VGG--16 model confirms a 1.25x performance gains is got by tuning the Winograd library. Integrated comparison results shows 1.02x to 1.40x, 1.14x to 2.17x, and 1.22x and 2.48x speedup is achieved over NNPACK, Arm NN, and FeatherCNN on the Kunpeng 920 beside few cases. Furthermore, problem size performance portability experiments with various convolution shapes shows that FastConv achieves 1.2x to 1.7x speedup and 2x to 22x speedup over NNPACK and ARM NN inference engine using Winograd on Kunpeng 920 . CPU performance portability evaluation on the VGG--16 show an average speedup over NNPACK of 1.42x, 1.21x, 1.26x, 1.37x, 2.26x, and 11.02x is observed on Kunpeng 920, Snapdragon 835, 855, 888, Apple M1, and AWS Graviton2, respectively.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Nanobody screening and machine learning guided identification of cross-variant anti-SARS-CoV-2 neutralizing heavy-chain only antibodies

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) continues to persist, demonstrating the risks posed by emerging infectious diseases to national security, public health, and the economy. Development of new vaccines and antibodies for emerging viral threats requires substantial resources and time, and traditional development platforms for vaccines and antibodies are often too slow to combat continuously evolving immunological escape variants, reducing their efficacy over time. Previously, we designed a next-generation synthetic humanized nanobody (Nb) phage display library and demonstrated that this library could be used to rapidly identify highly specific and potent neutralizing heavy chain-only antibodies (HCAbs) with prophylactic and therapeutic efficacy in vivo against the original SARS-CoV-2. In this study, we used a combination of high throughput screening and machine learning (ML) models to identify HCAbs with potent efficacy against SARS-CoV-2 viral variants of interest (VOIs) and concern (VOCs). To start, we screened our highly diverse Nb phage display library against several pre-Omicron VOI and VOC receptor binding domains (RBDs) to identify panels of cross-reactive HCAbs. Using HCAb affinity for SARS-CoV-2 VOI and VOCs (pre-Omicron variants) and model features from other published data, we were able to develop a ML model that successfully identified HCAbs with efficacy against Omicron variants, independent of our experimental biopanning workflow. This biopanning informed ML approach reduced the experimental screening burden by 78% to 90% for the Omicron BA.5 and Omicron BA.1 variants, respectively. The combined approach can be applied to other emerging viruses with pandemic potential to rapidly identify effective therapeutic antibodies against emerging variants.

Antibodies↗