Search NASA⌕ Search

SEARCH · Search NASA

Results for “sequence design”

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 199 records · Page 11
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Peptide nucleic acid Hoogsteen strand linker design for major groove recognition of DNA thymine bases

Sequence-specific targeting of double-stranded DNA and non-coding RNA via triple-helix-forming peptide nucleic acids (PNAs) has attracted considerable attention in therapeutic, diagnostic and nanotechnological fields. An E-base (3-oxo-2,3-dihydropyridazine), attached to the polyamide backbone of a PNA Hoogsteen strand by a side-chain linker molecule, is typically used in the hydrogen bond recognition of the 4-oxo group of thymine and uracil nucleic acid bases in the major groove. We report on the application of quantum chemical computational methods, in conjunction with spatial constraints derived from the experimental structure of a homopyrimidine PNA·DNA-PNA hetero-triplex, to investigate the influence of linker flexibility on binding interactions of the E-base with thymine and uracil bases in geometry-optimised model systems. Hydrogen bond formation between the N2 E-base atom and target pyrimidine base 4-oxo groups in model systems containing a β-alanine linker (J Am Chem Soc 119:11116, 1997) was found to incur significant internal strain energy and the potential disruption of intra-stand aromatic base stacking interactions in an oligomeric context. Additionally, in geometry-optimised model systems containing a 3-trans olefin linker (Bioorg Med Chem Lett 14:1551, 2004) the E-base swung out away from the target pyrimidine bases into the solvent. These findings are in qualitative agreement with calorimetric measurements in hybridisation experiments at T–A and U–A inversion sites. In contrast, calculations on a novel 2-cis olefin linker design indicate that it could permit simultaneous E-base hydrogen bonding with the thymine 4-oxo group, circumvention and solvent screening of the thymine 5-methyl group, and maintenance of triplex intra-stand base stacking interactions.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Moving the needle: Employing deep reinforcement learning to push the boundaries of coarse-grained vaccine models

Highly mutable infectious disease pathogens (hm-IDPs) such as HIV and influenza evolve faster than the human immune system can contain them, allowing them to circumvent traditional vaccination approaches and causing over one million deaths annually. Agent-based models can be used to simulate the complex interactions that occur between immune cells and hm-IDP-like proteins (antigens) during affinity maturation—the process by which antibodies evolve. Compared to existing experimental approaches, agent-based models offer a safe, low-cost, and rapid route to study the immune response to vaccines spanning a wide range of design variables. However, the highly stochastic nature of affinity maturation and vast sequence space of hm-IDPs render brute force searches intractable for exploring all pertinent vaccine design variables and the subset of immunization protocols encompassed therein. To address this challenge, we employed deep reinforcement learning to drive a recently developed agent-based model of affinity maturation to focus sampling on immunization protocols with greater potential to improve the chosen metrics of protection, namely the broadly neutralizing antibody (bnAb) titers or fraction of bnAbs produced. Using this approach, we were able to coarse-grain a wide range of vaccine design variables and explore the relevant design space. Our work offers new testable insights into how vaccines should be formulated to maximize protective immune responses to hm-IDPs and how they can be minimally tailored to account for major sources of heterogeneity in human immune responses and various socioeconomic factors. Our results indicate that the first 3 to 5 immunizations, depending on the metric of protection, should be specially tailored to achieve a robust protective immune response, but that beyond this point further immunizations require only subtle changes in formulation to sustain a durable bnAb response.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Practical choices for space-filling designs

Space-filling designs are now commonly used as a flexible model-free strategy for providing good coverage throughout an input space of interest for a variety of computer and physical experiment design scenarios. Some of the preliminary choices about how to frame the problem and which type of design to use can have a substantial impact on the success or failure of the experiment, and yet how to make these critical choices is often under-emphasized in the literature. In this paper, we explore several of the practical choices required by the experimenter and describe a sequence of steps to help create an ideal design that matches the goals and constraints of the experiment. These choices include the specification of the input space, the scaling of the variables, the degree of uniformity of the design points across the input space, and the space-filling characteristics. In addition, some new tools for defining the weights to implement non-uniform space-filling designs are provided. Finally, the methods are demonstrated with several illustrative examples and a real-world chemical engineering experiment for carbon capture.

42 ENGINEERING↗
DOE OSTI2020

Efficacy and Delivery of Novel FAST Agents for Coronaviruses

We proposed to test and develop advanced delivery for novel agents from our collaborators Facile Accelerated Specific Therapeutics (FAST) platform to reduce coronavirus replication. Sachi Bioworks Inc., Prof. Anushree Chatterjee, and Prof. Prashant Nagpal at the University of Colorado Boulder have developed a bioinformatics and synthesis pipeline to produce sequence specific theranostic agents (agents that can be therapies and/or diagnostics) that are inherently transported into the cytoplasm of mammalian host cells and sequence-specifically interfere in nucleic acid replication. The agent comprises a small nanoparticle (2-5 nm) chosen for ideal cellular transport and/or imaging conjugated to a short, synthetic DNA analog oligomer designed for binding to one or more target viral sequences. The sequence specific binding of the FAST agent to its target prevents nucleic acid replication due to its high affinity binding. While the small nanoparticle facilitates delivery in vitro, we plan to package the FAST agents into a larger nanoparticle (80-300 nm) for future in vivo delivery applications. Our team at Sandia has expertise encapsulating biomolecules including protein, DNA, and RNA into solid lipid nanoparticles (LNP) and lipid coated mesoporous silica nanoparticles (LC-MSN) and shown successful delivery in mouse models to multiple tissues. Our team focused on formulation parameters for FAST agents into lipid nanoparticles (LNP) and lipid coated mesoporous silica nanoparticles (LC-MSN) for enhanced delivery and/or efficacy and in vivo translation. We used lipid formulas that have been shown in literature to facility in vitro and more importantly, in vivo delivery. In our work discussed below, we successfully demonstrate loading and release of FAST agents on silica core and stable LC-MSN in a reasonable size range for in vivo testing.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Catalytic DNA Polymerization Can Be Expedited by Active Product Release**

Abstract The sequence‐specific hybridization of DNA facilitates its use as a building block for designer nanoscale structures and reaction networks that perform computations. However, the strong binding energy of Watson–Crick base pairing that underlies this specificity also causes the DNA dehybridization rate to depend sensitively on sequence length and temperature. This strong dependency imposes stringent constraints on the design of multi‐step DNA reactions. Here we show how an ATP‐dependent helicase, Rep‐X, can drive specific dehybridization reactions at rates independent of sequence length, removing the constraints of equilibrium on DNA hybridization and dehybridization. To illustrate how this new capacity can speed up designed DNA reaction networks, we show that Rep‐X extends the range of conditions where the primer exchange reaction, which catalytically adds a domain provided by a hairpin template to a DNA substrate, proceeds rapidly.

Moerman, Pepijn G.↗
DOE OSTI2022

Catalytic DNA Polymerization Can Be Expedited by Active Product Release**

Abstract The sequence‐specific hybridization of DNA facilitates its use as a building block for designer nanoscale structures and reaction networks that perform computations. However, the strong binding energy of Watson–Crick base pairing that underlies this specificity also causes the DNA dehybridization rate to depend sensitively on sequence length and temperature. This strong dependency imposes stringent constraints on the design of multi‐step DNA reactions. Here we show how an ATP‐dependent helicase, Rep‐X, can drive specific dehybridization reactions at rates independent of sequence length, removing the constraints of equilibrium on DNA hybridization and dehybridization. To illustrate how this new capacity can speed up designed DNA reaction networks, we show that Rep‐X extends the range of conditions where the primer exchange reaction, which catalytically adds a domain provided by a hairpin template to a DNA substrate, proceeds rapidly.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Identification and characterization of mono- and bifunctional galactan synthases in the pediatric pathogen Kingella kingae

The emerging pediatric pathogen Kingella kingae elaborates a lipopolysaccharide (LPS) that is extended with a galactofuranose homopolymer called galactan, which is a key virulence determinant that contributes to resistance to complement-mediated and neutrophil-mediated killing. Previous work has demonstrated that the pamABCDE locus is required for galactan synthesis. In this study, mutational studies suggested that the pamC gene product is a UDP-galactofuranose (Galf) transferase and is the galactan synthase. Analysis of genome sequence data revealed two distinct pamC alleles designated pamC1 and pamC2, which correlate with the two galactan structures in K. kingae. Examination of isogenic mutants expressing either pamC1 or pamC2 demonstrated that the pamC alleles are the determinants of galactan structure. Experiments with recombinant PamC1 and PamC2 in vitro established that these proteins are galactan synthases capable of extending synthetic Galf disaccharide acceptors in the presence of UDP-Galf. Homology analysis identified critical amino acids that are essential for PamC1 and PamC2 enzymatic activity both in vitro and in K. kingae. Structural analysis of the in vitro-modified synthetic acceptors implicated PamC1 as a monofunctional enzyme capable of generating a β-(1 → 5) Galf linkage and PamC2 as a bifunctional enzyme capable of generating β-(1 → 3) and β-(1 → 6) Galf linkages. This study advances our understanding of the GT2 family of UDP-galactofuranosyltransferases.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

PCP consensus protein/peptide alphavirus antigens stimulate broad spectrum neutralizing antibodies

Vaccines based on proteins and peptides may be safer and if calculated based on many sequences, more broad-spectrum than those designed based on single strains. Physicochemical Property Consensus (PCP con ) alphavirus (AV) antigens from the B-domain of the E2 envelope protein were designed, synthesized recombinantly and shown to be immunogenic (i.e. sera after inoculation detected the antigen in dotspots and ELISA). Antibodies in sera after inoculation with B-region antigens based on individual AV species (eastern or Venezuelan equine encephalitis (EEEV con , VEEV con ), or chikungunya (CHIKV con ) bound only their cognate protein, while those designed against multiple species (Mosaik con and EVC con ) recognized all three serotype specific antigens. The VEEV con and EEEV con sera only showed antiviral activity against their related strains (in plaque reduction neutralization assays (PRNT 50/80 ). Peptides designed to surface exposed areas of the E2-A-domain of CHIKV con were added to CHIKV con inocula to provide anti-CHIKV antibodies. EVC con , based on three different alphavirus species, combined with E2-A-domain peptides from AllAV con , a PCPcon of 24 diverse AV, generated broad spectrum, antiviral antibodies against VEEV, EEEV and CHIKV, AV with less than 35% amino acid identity to each other (>65% diversity). This is a promising start to a molecularly defined vaccine against all AV. A further study with these antigens can illuminate what areas are most important for a robust immune response, resistant to mutations in rapidly evolving viruses. The validated computational methods can also be used to design broad spectrum antigens against many other pathogen families.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

A FAIR-compliant parts catalogue for genome engineering and expression control in Saccharomyces cerevisiae

The synthetic biology toolkit for baker's yeast, Saccharomyces cerevisiae, includes extensive genome engineering toolkits and parts repositories. However, with the increasing complexity of engineering tasks and versatile applications of this model eukaryote, there is a continued interest to expand and diversify the rational engineering capabilities in this chassis by FAIR (findable, accessible, interoperable, and reproducible) compliance. In this study, we designed and characterised 41 synthetic guide RNA sequences to expand the CRISPR-based genome engineering capabilities for easy and efficient replacement of genomically encoded elements. Moreover, we characterize in high temporal resolution 20 native promoters and 18 terminators using fluorescein and LUDOX CL-X as references for GFP expression and OD600 measurements, respectively. Additionally, all data and reported analysis is provided in a publicly accessible jupyter notebook providing a tool for researchers with low-coding skills to further explore the generated data as well as a template for researchers to write their own scripts. We expect the data, parts, and databases associated with this study to support a FAIR-compliant resource for further advancing the engineering of yeasts.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Native Chemical Ligation of Peptoid Oligomers

Bioorganic chemists are inspired by natural biopolymers to design peptidomimetic oligomers that can exhibit sequence-structure-function relationships. Biomimetic polymers can be synthesized to incorporate a specific sequence of nonbiological monomer units using a variety of iterative solution-phase or solid-phase reaction schemes. These protocols generally provide access to a vast diversity of oligomeric compounds but are limited with respect to their ability to attain protein-like chain lengths. This constraint can preclude access to sequence-defined synthetic macromolecules with sufficient sizes required to exhibit tertiary structure and other protein-mimetic attributes. In contrast, peptide chemists have overcome this limitation by developing convergent synthetic methods, such as native chemical ligation, to join individual, smaller peptide chains together to make larger peptides or full proteins. A similar convergent approach is needed to establish efficient synthetic routes to non-natural sequence-defined macromolecules. Herein, we adapt the peptide native chemical ligation method to peptoid oligomers, demonstrating how short chains can be conjoined to create sequence-defined peptoid macromolecules. Nanosheet-forming peptoid polymers with distinct surface loop display domains were generated by sequential ligation of several discrete fragments. This method provides a reliable convergent ligation route for sequence-defined polypeptoids that results in a native amide bond joining the fragments. We envision that this strategy will be useful in synthesizing peptoid-based proteomimetics that incorporate diverse chemical features.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Redshift evolution of the H2/H i mass ratio in galaxies

ABSTRACT In this paper, we present an attempt to estimate the redshift evolution of the molecular to neutral gas mass ratio within galaxies (at fixed stellar mass). For a sample of five nearby grand design spirals located on the main-sequence (MS) of star-forming galaxies, we exploit maps at 500 pc resolution of stellar mass and star formation rate (M⋆ and SFR). For the same cells, we also have estimates of the neutral (MH i) and molecular ($M_{\rm H_2}$) gas masses. To compute the redshift evolution, we exploit two relations: (i) one between the molecular-to-neutral mass ratio and the total gas mass (Mgas), whose scatter shows a strong dependence with the distance from the spatially resolved MS, and (ii) the one between $\log (M_{\rm {H_2}}/M_{\star })$ and log (MH i/M⋆). For both methods, we and that $M_{\rm H_2}$/MH i within the optical radius slightly decreases with redshift, contrary to common expectations of galaxies becoming progressively more dominated by molecular hydrogen at high redshifts. We discuss possible implications of this trend on our understanding of the internal working of high-redshift galaxies.

Morselli, Laura (ORCID:0000000307532571)↗
DOE OSTI2022

Efficacy of Cathelicidin-Mimetic Antimicrobial Peptoids against Staphylococcus aureus

Staphylococcus aureus is one of the most common pathogens associated with infection in wounds. The current standard of care uses a combination of disinfection and drainage followed by conventional antibiotics such as methicillin. Methicillin and vancomycin resistance has rendered these treatments ineffective, often causing the reemergence of infection. This study examines the use of antimicrobial peptoids (sequence-specific poly-N-substituted glycines) designed to mimic naturally occurring cationic, amphipathic host defense peptides, as an alternative to conventional antibiotics. These peptoids also show efficient and fast (<30 min) killing of methicillin-susceptible S. aureus (MSSA) and methicillin-resistant S. aureus (MRSA) at low micromolar concentrations without having apparent cytotoxic side effects in vivo. Additionally, these novel peptoids show excellent efficacy against biofilm formation and detachment for both MSSA and MRSA. In comparison, conventional antibiotics were unable to detach or prevent formation of biofilms. One cationic 12mer, Peptoid 1, shows great promise, as it could prevent formation of and detach biofilms at concentrations as low as 1.6 μM. The use of a bioluminescent S. aureus murine incision wound model demonstrated clearance of infection in peptoid-treated mice within 8 days, conveying another advantage these peptoids have over conventional antibiotics. These results provide clear evidence of the potential for antimicrobial peptoids for the treatment of S. aureus wound infections.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

OpenStudio®-MCP [SWR-26-035]

OpenStudio®-MCP is a Model Context Protocol (MCP) server that lets AI assistants perform building energy modeling through natural language. Rather than requiring users to learn the OpenStudio® SDK, EnergyPlus® scripting, or Ruby/Python automation, the server translates conversational requests into sequences of tool calls that create models, design HVAC systems, run simulations, and extract results — all within a single chat session. The server's 124 tools are organized into a skills architecture where each skill encapsulates a domain of building energy modeling (envelope, HVAC, loads, weather, simulation, results) behind typed, LLM-friendly interfaces. High-leverage operations like applying ASHRAE 90.1 baseline systems or generating standards-compliant typical buildings are exposed as single tool calls that internally wire dozens of OpenStudio® objects. Bundled measures from ComStock™ and Openstudio® -common-measures-gem are wrapped with dedicated tools and typed arguments rather than exposed through a generic measure interface, so AI models get consistent, error-resistant recipes without needing to discover measure arguments at runtime. A key design decision is structured results extraction: six SQL-based tools return surgical ~300–1,000 token responses (end-use breakdowns, envelope summaries, HVAC sizing, timeseries data) instead of requiring the AI to parse ~100K-token raw HTML reports, making iterative design exploration practical within context window limits. The codebase is designed as a reference implementation — explicit, well-commented, and modular — so that other simulation engines (EnergyPlus® standalone, TRNSYS, DOE-2) can use it as a template for building their own MCP servers.

Ball, Brian [National Laboratory of the Rockies (N↗
DOE OSTI2025

Fractionation of Filamentous Algae from Mixed Biofilms

Filamentous algae, which grow in long, hair-like filaments within biofilms, play a crucial role in wastewater treatment due to their ability to produce significant biomass and their resistance to predation compared to traditional microalgal treatments. These algae can effectively uptake and utilize pollutants, particularly excessive nitrogen (ammonia, nitrate, nitrite) and phosphorus (phosphate), making filamentous algae valuable for wastewater treatment, as well as bioethanol and biodiesel production due to high lipid productions. However, each algal species possesses different capacities, necessitating a thorough genetic identification and understanding of each community. A major challenge in accurately assessing these communities is the lack of coverage in large sequencing databases which can lead to misrepresentation of the true composition and abundance of organisms and overall sequencing bias. To address this, I evaluated chemical and physical techniques for separating filamentous algae from mixed biofilms to achieve clean genetic sequencing results. I employed pH washing (0.001M HCl, 0.001M HCl, DiH2O, 0.0001M HCl, 0.001M HCl) for chemical treatment, followed by physical separation through centrifugation (5000rpm, 6500rpm) or filtration (2mm, 250um, 75um). The most successful method was deionized water washing, which yielded clear differences across stacked filters; the 2mm filtrate showed high levels of filamentous algae, with microalgae eluting in the 75um filtrate or remaining within agglutinations of algae larger filters. Base washing eluted the highest concentrations of microalgae, with larger filter sizes retaining more filamentous algae, indicating the breakdown of extracellular polymeric substances (EPS). Our downstream plans include sending the high-throughput next-generation sequencing to confirm the purity and ratios of filamentous and non-filamentous algae, as well as bacteria present, thereby validating the success of our treatments. Potential applications include creating community-based fractions for analysis, refining current sequencing data with clearer isolations, and generating designer biofilms to enhance our understanding of community interactions.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Estimating Residual Gas Beam Loss in the BNL AGS Booster:Model, Numerical Tools, and Implications for a Polarimeter Chamber

Beamloss from residual gas scattering is calculated for a storage ring of circumference C = 200m, covering a linear magnetic rigidity ramp from B ρ inj = 1.25Tm to B ρ ext = 9.50Tm over tramp = 1.00s. Three loss mechanisms are considered: nuclear inelastic reactions, single Coulomb scattering beyond the acceptance angle of the machine, and charge-state exchange for heavy ions. Calculations are performed for protons and Au 32+ ions, the latter being the charge state delivered by the EBIS pre-injector into the BNL AGS Booster. At a baseline vacuum of P 0 = 5 × 10 −11 mbar the integrated fractional beam loss over the full ramp is negligible for protons (< 0.001%) and 33.1% for Au 32+ . The impact of a 1m polarimeter section at elevated pressure P s = 5 × 10 −9 mbar is evaluated: the proton loss remains negligible, while the Au 32+ loss rises to 45.1%, establishing the vacuum requirement for the polarimeter chamber. In the present compact approximate model, Au 32+ losses are dominated by charge exchange at injection, while proton losses are dominated by single Coulomb scattering. The design, outgassing budget, pumping strategy, and commissioning sequence for a UHV polarimeter chamber housing silicon detectors and ASICs are described in detail.

43 PARTICLE ACCELERATORS↗
DOE OSTI2026

Assessing Options for Particulate Measurement Tools to Advance Wood Heaters (CRADA Final Report)

For approximately 40 years, particulate matter (PM) from residential wood heaters has been certified in North America using an EPA dilution tunnel with a filter-based gravimetric measurement. While reasonably repeatable, this approach reports only a single value for the entire burn sequence and does not provide information on PM at different points within the sequence, which would support improved wood heater designs. PM monitors that report with high time resolution offer a potential solution, but their performance for various types of heaters and across heater operating conditions has not been extensively evaluated. In this 18-month collaborative project, the Hearth, Patio & Barbecue Association (HPBA) and Lawrence Berkeley National Laboratory (LBNL) tested five continuous PM instruments spanning three measurement principles—light-scattering, inertial mass (TEOM), and triboelectric—against gravimetric and size-resolved references on three residential heater types (cordwood, catalytic cordwood, and pellet) across startup, high-burn, and low-burn operating phases.

42 ENGINEERING↗
DOE OSTI2022

Disc dichotomy signature in the vertical distribution of [Mg/Fe] and the delayed gas infall scenario

Context. Analysis of the Apache Point Observatory Galactic Evolution Experiment project (APOGEE) data suggests the existence of a clear distinction between two sequences of disc stars in the [α/Fe] versus [Fe/H] abundance ratio space, known as the high- and low-α sequence, respectively. This dichotomy also emerges from an analysis of the vertical distribution of the [α/Fe] abundance ratio. Aims. We aim to test whether the revised two-infall chemical evolution models designed to reproduce the low- and high-α sequences in the [α/Fe] versus [Fe/H] ratios in the solar neighbourhood are also capable of predicting the disc bimodality observed in the vertical distribution of [Mg/Fe] in APOGEE DR16 data. Methods. Along with the chemical composition of the simple stellar populations born at different Galactic times predicted by our reference chemical evolution models in the solar vicinity, we provide their maximum vertical height above the Galactic plane |zmax| computed assuming the relation between the vertical action and stellar age in APOGEE thin-disc stars. Result. The vertical distribution of the [Mg/Fe] abundance ratio predicted by the reference chemical evolution models is in agreement with that observed when combining the APOGEE DR16 data (chemical abundances) with the astroNN catalogue (stellar ages, orbital parameters) for stars younger than 8 Gyr (only low-α sequence stars). Including the high-α disc component, the dichotomy in the vertical [Mg/Fe] abundance distribution is reproduced considering the observational cut in the Galactic height of |z|< 2 kpc. However, our model predicts an overly flat (almost constant) growth of the maximum vertical height |zmax| quantity as a function of [Mg/Fe] for high-α objects in contrast with the median values from APOGEE data. Possible explanations for such a tension are that: (i) the APOGEE sample with |z|< 2 kpc is more likely than ours to be contaminated by halo stars, causing the median values to be kinematically hotter, and (ii) external perturbations – such as minor mergers – that the Milky Way experienced in the past could have heated up the disc, and the heating of the orbits cannot be modeled by only scattering processes. Assuming a disc dissection based on chemistry for APOGEE-DR16 stars (|z|< 2 kpc), the observed |zmax| distributions for high-α and low-α sequences are in good agreement with our model predictions if we consider the errors in the vertical action estimates in the calculation. Moreover, a better agreement between predicted and observed stellar distributions at different Galactic vertical heights is achieved if asteroseismic ages are included as a constraint in the best-fit model calculations. Conclusions. The signature of a delayed gas infall episode, which gives rise to a hiatus in the star formation history of the Galaxy, are imprinted both in the [Mg/Fe] versus [Fe/H] relation and in vertical distribution of [Mg/Fe] abundances in the solar vicinity.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Large language models generate functional protein sequences across diverse families

Deep-learning language models have shown promise in various biotechnological applications, including protein design and engineering. Here, in this paper, we describe ProGen, a language model that can generate protein sequences with a predictable function across large protein families, akin to generating grammatically and semantically correct natural language sentences on diverse topics. The model was trained on 280 million protein sequences from >19,000 families and is augmented with control tags specifying protein properties. ProGen can be further fine-tuned to curated sequences and tags to improve controllable generation performance of proteins from families with sufficient homologous samples. Artificial proteins fine-tuned to five distinct lysozyme families showed similar catalytic efficiencies as natural lysozymes, with sequence identity to natural proteins as low as 31.4%. ProGen is readily adapted to diverse protein families, as we demonstrate with chorismate mutase and malate dehydrogenase.

59 BASIC BIOLOGICAL SCIENCES↗