Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Homology modelling”

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 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Mechanisms of Three-Dimensional Solid-Phase Epitaxial Crystallization of Strontium Titanate

Strontium titanate (SrTiO 3 , STO) is a complex metal oxide with a cubic perovskite crystal structure. Due to its easily described and understood crystal structure in the cubic phase, STO is an ideal model system for exploring the mechanistic details of solid-phase epitaxy (SPE) in complex oxides. SPE is a crystallization approach that aims to guide crystal growth at low homologous temperatures to achieve targeted microstructures. Beyond planar thin films, SPE can also exploit the addition of a chemically inert, noncrystallizing, amorphous obstacle in the path of crystallization to generate complex three-dimensional structures. The introduction of this mask fundamentally alters the SPE process, inducing a transition from two- to three-dimensional geometries and from vertical to lateral crystal growth under the influence of the crystal/mask/amorphous boundary. Using a combination of molecular dynamics simulations and experiments, we identify several unique phenomena in the nanoscale growth behaviors in both conventional (unmasked) and masked SPE. Examining conventional SPE of STO, we find that crystallization at the interface is strongly correlated to, and potentially driven by, density fluctuations in the region of the amorphous STO near the crystalline/amorphous interface with a strong facet dependence. In the masked case, we find that the crystalline growth front becomes nonplanar near contact with the mask. We also observe a minimum vertical growth requirement prior to lateral crystallization. Both phenomena depend on the relative bulk and interfacial free energies of the three-phase (crystal/mask/amorphous) system.

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

De novo design of transmembrane β barrels

Transmembrane β-barrel proteins (TMBs) are of great interest for single-molecule analytical technologies because they can spontaneously fold and insert into membranes and form stable pores, but the range of pore properties that can be achieved by repurposing natural TMBs is limited. We leverage the power of de novo computational design coupled with a “hypothesis, design, and test” approach to determine TMB design principles, notably, the importance of negative design to slow β-sheet assembly. We design new eight-stranded TMBs, with no homology to known TMBs, that insert and fold reversibly into synthetic lipid membranes and have nuclear magnetic resonance and x-ray crystal structures very similar to the computational models. These advances should enable the custom design of pores for a wide range of applications.

Vorobieva, Anastassia A.↗
DOE OSTI2023

Deoptimization of FMDV P1 Region Results in Robust Serotype-Independent Viral Attenuation

Foot-and-mouth disease (FMD), caused by the FMD virus (FMDV), is a highly contagious disease of cloven-hoofed livestock that can have severe economic impacts. Control and prevention strategies, including the development of improved vaccines, are urgently needed to effectively control FMD outbreaks in endemic settings. Previously, we employed two distinct strategies (codon pair bias deoptimization (CPD) and codon bias deoptimization (CD)) to deoptimize various regions of the FMDV serotype A subtype A12 genome, which resulted in the development of an attenuated virus in vitro and in vivo, inducing varying levels of humoral responses. In the current study, we examined the versatility of the system by using CPD applied to the P1 capsid coding region of FMDV serotype A subtype, A24, and another serotype, Asia1. Viruses carrying recoded P1 (A24-P1Deopt or Asia1-P1Deopt) exhibited different degrees of attenuation (i.e., delayed viral growth kinetics and replication) in cultured cells. Studies in vivo using a mouse model of FMD demonstrated that inoculation with the A24-P1Deopt and Asia1-P1Deopt strains elicited a strong humoral immune response capable of offering protection against challenge with homologous wildtype (WT) viruses. However, different results were obtained in pigs. While clear attenuation was detected for both the A24-P1Deopt and Asia1-P1Deopt strains, only a limited induction of adaptive immunity and protection against challenge was detected, depending on the inoculated dose and serotype deoptimized. Our work demonstrates that while CPD of the P1 coding region attenuates viral strains of multiple FMDV serotypes/subtypes, a thorough assessment of virulence and induction of adaptive immunity in the natural host is required in each case in order to finely adjust the degree of deoptimization required for attenuation without affecting the induction of protective adaptive immune responses.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Transforming our understanding of chloroplast-associated genes through comprehensive characterization of protein localizations and protein-protein interactions

Bioenergy crops are a renewable source of fuels and are a critical base for building a carbon-neutral economy. Rational engineering of bioenergy crops has the potential to enhance the yields. However, our ability to engineer plants is limited because the functions of most genes remain unknown. Systematic characterization of gene function in plants thus has the potential to greatly accelerate bioenergy research. Here, we focus on the chloroplast, an underexplored energy-producing organelle that is a hallmark of plants. The chloroplast is one of the promising targets of biofuel crop engineering efforts because of its central role in photosynthesis, metabolism, and intracellular signaling. However, the protein composition of the chloroplast and the functions of most of its proteins remain poorly characterized. At the core of this project, we sought to comprehensively determine the localization of chloroplast-associated proteins and generate a spatially defined protein-protein interaction network for chloroplast. For this purpose, we used the leading model alga Chlamydomonas reinhardtii, which greatly increased experimental speed and throughput. We illustrated the value of our findings to land plants by determining the localization of Arabidopsis thaliana land plant homologs of the Chlamydomonas proteins. Altogether, we were successful in determining the localization of 1,034 chloroplast-associated proteins in Chlamydomonas. The localizations provide numerous insights into the spatial organization of chloroplasts and how they function to support photosynthesis. The localization patterns of distinct proteins revealed new chloroplast structures and revealed new spatial organization inside the chloroplast. We also identified new components of known chloroplast structures, such as the chloroplast envelope, nucleoid, plastoglobuli, and pyrenoid. We identified these new components by investigating the interacting partners of known proteins. Many proteins localized in both the chloroplast and other cellular structures, thereby hinting at new functions and communication between cellular structures. We also applied machine learning on the atlas to generate predictions for the location of all of the proteins in Chlamydomonas. This enabled us to assign putative functions to many uncharacterized proteins based on their cellular location. Altogether, this research establishes a rich resource that opens new avenues of investigation and guides future work in deciphering and manipulating chloroplast function. Next, we developed an extensive protein-protein interaction network for the chloroplast by performing affinity purification-mass spectrometry on ~1,150 tagged chloroplast-associated proteins, the first such large-scale study in any photosynthetic organism. This dataset reveals 4,694 high-confidence protein-protein interactions, offering insights into the functions of thousands of conserved poorly-characterized chloroplast proteins. This systematic identification of protein-protein interactions in the chloroplast also provides multiple exciting new research directions and a detailed blueprint of the chloroplast's operation. This research lays the groundwork to decipher the inner workings of the chloroplast, the cell structure at the heart of photosynthesis. The spatial atlas and protein-protein interactions reveal chloroplast organizational features that would not have been accessible with traditional approaches. The localization mapping, insights into the function, and research materials generated further provide a rich resource for the research community to advance the understanding of how the chloroplast is organized to enable engineering of enhanced photosynthetic organisms.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

FHOD formin and SRF promote post-embryonic striated muscle growth through separate pathways in C. elegans

Highlights: • Reduced SRF or formin FHOD-1 activity similarly inhibit muscle growth in worms. • Formin does not promote worm SRF-dependent muscle gene transcription. • Proteasome-driven myosin proteolysis is increased in worm muscle lacking formin. Previous work with cultured cells has shown transcription of muscle genes by serum response factor (SRF) can be stimulated by actin polymerization driven by proteins of the formin family. However, it is not clear if endogenous formins similarly promote SRF-dependent transcription during muscle development in vivo. We tested whether formin activity promotes SRF-dependent transcription in striated muscle in the simple animal model, Caenorhabditis elegans. Our lab has shown FHOD-1 is the only formin that directly promotes sarcomere formation in the worm's striated muscle. We show here FHOD-1 and SRF homolog UNC-120 both support muscle growth and also muscle myosin II heavy chain A expression. However, while a hypomorphic unc-120 allele blunts expression of a set of striated muscle genes, these genes are largely upregulated or unchanged by absence of FHOD-1. Instead, pharmacological inhibition of the proteasome restores myosin protein levels in worms lacking FHOD-1, suggesting elevated proteolysis accounts for their myosin deficit. Interestingly, proteasome inhibition does not restore normal muscle growth to fhod-1(Δ) mutants, suggesting formin contributes to muscle growth by some alternative mechanism. Overall, we find SRF does not depend on formin to promote muscle gene transcription in a simple in vivo system.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Structures of two novel crystal forms of Aspergillus oryzae alpha amylase (taka-amylase)

The structures of Aspergillus oryzae α-amylase were determined in a tetragonal crystal, having one molecule as asymmetric unit, and a monoclinic crystal with two molecules as asymmetric unit. Both crystal forms were obtained from trace contaminants of an old commercial lipase preparation. Structures were determined and refined to 1.65 Å and 1.43 Å resolution respectively. The latter crystal has a non-crystallographic (NCS) twofold axis within the asymmetric unit. Glycosylation at Asn197 is evident, and in the tetragonal crystal can be seen to include three, partially disordered sugar residues following the initial N-acetyl glucosamine (NAG). Superposition of the tetragonal crystal model on the α-amylases from Bacillus subtilis (PDB:1BAG), pig pancreas (PDB:3L2L), and barley (PDB:1AMY), show a high degree of coincidence, particularly for the (β/α)8-barrel domains, and especially within the active site. Using this structural agreement between amylases, we extrapolated the binding model of a six residue, limit dextrin found in pig pancreas α-amylase to the A. oryzae enzyme model, which predicts substrate interacting amino acid residues.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Polyketide synthase-based controlled synthesis of polycyclopropanated fuel molecules

Reducing carbon emissions from aviation and long-distance transportation sectors requires the development of sustainable biofuels with suitable energy density, freezing point, and other physical properties. We previously demonstrated biological production of high energy polycyclopropanated fatty acids (POP-FAs, class I) using an iterative polyketide synthase (iPKS) pathway in a Streptomyces host. Here, we used a computational model of fuel properties to identify chain length and cyclopropanation control as critical steps to engineer this iPKS for biofuel applications. We next explored the natural diversity of POP biosynthesis by investigating homologous pathways. Then, by in vivo gene exchange, we determined cyclopropanase (CP) catalysis to be key for POP-FA engineering. Leveraging both natural and engineered pathway product diversity, we demonstrate targeted production of improved POP-FAs, namely shortened POP-FAs with predicted superior freezing point properties for aviation, as well as fully cyclopropane-saturated POP-FAs which should have superior energy-density. These precise and controllable modifications to POP-FA structure open the door for bioproduction of designer POP fuels.

Yin, Kevin↗
DOE OSTI2022

Phage K gp102 Drives Temperature-Sensitive Antibacterial Activity on USA300 MRSA

There is widespread interest in using obligately lytic bacteriophages (“phages”) to treat human bacterial infections. Among Staphylococcus aureus infections, the USA300 lineage is a frequent cause of invasive disease. We observed that phage K, a model S. aureus myophage, exhibits temperature-sensitive growth on USA300 strains, with the wild-type phage providing poorer growth suppression in broth and forming smaller and fainter plaques at 37 °C vs. 30 °C. We isolated 65 mutants of phage K that had improved plaquing characteristics at 37 °C when compared to the parental phage. In all 65 mutants, this phenotype was attributable to loss-of-function (LoF) mutations in gp102, which encodes a protein of unknown function that has homologs only among the Herelleviridae (SPO1-like myophages infecting gram-positive bacteria). Additional experiments with representative mutants consistently showed that the temperature-sensitive plaque phenotype was specific to USA300 MRSA strains and that Gp102 disruption was correlated with improved suppression of bacterial growth in broth and improved antibacterial activity in a mouse model of upper respiratory tract infection. The same genotype and in vitro phenotypes could be replicated in close relatives of phage K. Gp102 disruption did not have a detectable effect on adsorption but did delay cell culture lysis relative to wild-type under permissive infection conditions, suggesting that gp102 conservation might be maintained by selective pressure for more rapid replication. Expression of gp102 on a plasmid was toxic to both an MSSA and a USA300 MRSA strain. Molecular modeling predicts a protein with two helix-turn-helix domains that displays some similarity to DNA-binding proteins such as transcription factors. While its function remains unclear, gp102 is a conserved gene that is important to the infection process of Kayvirus phages, and it appears that the manner in which USA300 strains defend against them at 37 °C can be overcome by gp102 LoF mutations.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

PlasmidHostFinder: Prediction of Plasmid Hosts Using Random Forest

Plasmids play a major role facilitating the spread of antimicrobial resistance between bacteria. Understanding the host range and dissemination trajectories of plasmids is critical for surveillance and prevention of antimicrobial resistance. Identification of plasmid host ranges could be improved using automated pattern detection methods compared to homology-based methods due to the diversity and genetic plasticity of plasmids. In this study, we developed a method for predicting the host range of plasmids using machine learning—specifically, random forests. We trained the models with 8,519 plasmids from 359 different bacterial species per taxonomic level; the models achieved Matthews correlation coefficients of 0.662 and 0.867 at the species and order levels, respectively. Our results suggest that despite the diverse nature and genetic plasticity of plasmids, our random forest model can accurately distinguish between plasmid hosts. This tool is available online through the Center for Genomic Epidemiology (https://cge.cbs.dtu.dk/services/PlasmidHostFinder/).

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Dynamic Profiling of β-Coronavirus 3CL Mpro Protease Ligand-Binding Sites

β-coronavirus (CoVs) alone has been responsible for three major global outbreaks in the 21st century. The current crisis has led to an urgent requirement to develop therapeutics. Even though a number of vaccines are available, alternative strategies targeting essential viral components are required as a backup against the emergence of lethal viral variants. One such target is the main protease (Mpro) that plays an indispensable role in viral replication. The availability of over 270 Mpro X-ray structures in complex with inhibitors provides unique insights into ligand–protein interactions. Herein, we provide a comprehensive comparison of all nonredundant ligand-binding sites available for SARS-CoV2, SARS-CoV, and MERS-CoV Mpro. Extensive adaptive sampling has been used to investigate structural conservation of ligand-binding sites using Markov state models (MSMs) and compare conformational dynamics employing convolutional variational auto-encoder-based deep learning. Our results indicate that not all ligand-binding sites are dynamically conserved despite high sequence and structural conservation across β-CoV homologs. This highlights the complexity in targeting all three Mpro enzymes with a single pan inhibitor.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Potential virus-mediated nitrogen cycling in oxygen-depleted oceanic waters

Viruses play an important role in the ecology and biogeochemistry of marine ecosystems. Beyond mortality and gene transfer, viruses can reprogram microbial metabolism during infection by expressing auxiliary metabolic genes (AMGs) involved in photosynthesis, central carbon metabolism, and nutrient cycling. While previous studies have focused on AMG diversity in the sunlit and dark ocean, less is known about the role of viruses in shaping metabolic networks along redox gradients associated with marine oxygen minimum zones (OMZs). Here, we analyzed relatively quantitative viral metagenomic datasets that profiled the oxygen gradient across Eastern Tropical South Pacific (ETSP) OMZ waters, assessing whether OMZ viruses might impact nitrogen (N) cycling via AMGs. Identified viral genomes encoded six N-cycle AMGs associated with denitrification, nitrification, assimilatory nitrate reduction, and nitrite transport. The majority of these AMGs (80%) were identified in T4-like Myoviridae phages, predicted to infect Cyanobacteria and Proteobacteria, or in unclassified archaeal viruses predicted to infect Thaumarchaeota. Four AMGs were exclusive to anoxic waters and had distributions that paralleled homologous microbial genes. Together, these findings suggest viruses modulate N-cycling processes within the ETSP OMZ and may contribute to nitrogen loss throughout the global oceans thus providing a baseline for their inclusion in the ecosystem and geochemical models.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

cTULIP: application of a human-based RNA-seq primary tumor classification tool for cross-species primary tumor classification in canine

The domestic dog, Canis familiaris, is quickly gaining traction as an advantageous model for use in the study of cancer, one of the leading causes of death worldwide. Naturally occurring canine cancers share clinical, histological, and molecular characteristics with the corresponding human diseases. In this study, we take a deep-learning approach to test how similar the gene expression profile of canine glioma and bladder cancer (BLCA) tumors are to the corresponding human tumors. We likewise develop a tool for identifying misclassified or outlier samples in large canine oncological datasets, analogous to that which was developed for human datasets. We test a number of machine learning algorithms and found that a convolutional neural network outperformed logistic regression and random forest approaches. We use a recently developed RNA-seq-based convolutional neural network, TULIP, to test the robustness of a human-data-trained primary tumor classification tool on cross-species primary tumor prediction. Our study ultimately highlights the molecular similarities between canine and human BLCA and glioma tumors, showing that protein-coding one-to-one homologs shared between humans and canines, are sufficient to distinguish between BLCA and gliomas. The results of this study indicate that using protein-coding one-to-one homologs as the features in the input layer of TULIP performs good primary tumor prediction in both humans and canines. Furthermore, our analysis shows that our selected features also contain the majority of features with known clinical relevance in BLCA and gliomas. Our success in using a human-data-trained model for cross-species primary tumor prediction also sheds light on the conservation of oncological pathways in humans and canines, further underscoring the importance of the canine model system in the study of human disease.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Reductive amination cascades in cell‐free and resting whole cell formats for valorization of lignin deconstruction products

Abstract The selective introduction of amine groups within deconstruction products of lignin could provide an avenue for valorizing waste biomass while achieving a green synthesis of industrially relevant building blocks from sustainable sources. Here, we built and characterized enzyme cascades that create aldehydes and subsequently primary amines from diverse lignin‐derived carboxylic acids using a carboxylic acid reductase (CAR) and an ω‐transaminase (TA). Unlike previous studies that have paired CAR and TA enzymes, here we examine multiple homologs of each of these enzymes and a broader set of candidate substrates. In addition, we compare the performance of these systems in cell‐free and resting whole‐cell biocatalysis formats using the conversion of vanillate to vanillyl amine as model chemistry. We also demonstrate that resting whole cells can be recycled for multiple batch reactions. We used the knowledge gained from this study to produce several amines from carboxylic acid precursors using one‐pot biocatalytic reactions, several of which we report for the first time. These results expand our knowledge of these industrially relevant enzyme families to new substrates and contexts for environmentally friendly and potentially low‐cost synthesis of diverse aryl aldehydes and amines.

Biotechnology & Applied Microbiology↗
DOE OSTI2026

Assessing the Impacts of Conformational Fluxionality on Copper(II/I) Electron Transfer Self-Exchange

Typical Cu(II/I) complexes exhibit hallmark structural changes during their electron transfer (ET) reactions that result from the (pseudo) Jahn-Teller distortions and changes in polarizability inherent to their d 10 / d 9 configurations. Given that such structural changes incur large reorganization energy penalties, slow the rates of ET are characteristic of these compounds. Notwithstanding, we recently reported a set of Cu(II/I) complexes that undergo significant and well-defined structural changes during their redox reactions yet exhibit rapid (> 10 5 M ‒1 s ‒1 ) ET self-exchange rate constants ( k 11 ). To explain these results we proposed a pre-equilibrium model in which inherent conformational fluxionality in one of the two oxidation states provides access to pathways involving lower reorganization energies during the ET event. Herein we report our results testing this hypothesis through the preparation and study of a homologous series of compounds exhibiting varying extents of con-formational fluxionality in the Cu(I) state. We characterize these compounds electrochemically, structurally, and by varia-ble temperature NMR spectroscopy to provide experimental evidence for increase fluxionality across the series. We then correlate the trend with increasing k 11 through NMR linewidth broadening experiments, further taking care to define the impacts of solvent impurities therein. Finally, the nature of the conformational rearrangements, and their relation to increased k 11 are explored computationally to reveal differential Boltzmann populations of conformers across the series. The cumulative results of these studies support a previously underappreciated strategy for overcoming barriers to slow ET kinetics: via the incorporation of conformational fluxionality.

Singh, Aditi↗
DOE OSTI2023

Two-Dimensional Energy Histograms as Features for Machine Learning to Predict Adsorption in Diverse Nanoporous Materials

A major obstacle for machine learning (ML) in chemical science is the lack of physically informed feature representations that provide both accurate prediction and easy interpretability of the ML model. In this work, we describe adsorption systems using novel two-dimensional energy histogram (2D-EH) features, which are obtained from the probe-adsorbent energies and energy gradients at grid points located throughout the adsorbent. The 2D-EH features encode both energetic and structural information of the material and lead to highly accurate ML models (coefficient of determination R2 ~ 0.94–0.99) for predicting single-component adsorption capacity in metal–organic frameworks (MOFs). Here, we consider the adsorption of spherical molecules (Kr and Xe), linear alkanes with a wide range of aspect ratios (ethane, propane, n-butane, and n-hexane), and a branched alkane (2,2-dimethylbutane) over a wide range of temperatures and pressures. The interpretable 2D-EH features enable the ML model to learn the basic physics of adsorption in pores from the training data. We show that these MOF-data-trained ML models are transferrable to different families of amorphous nanoporous materials. We also identify several adsorption systems where capillary condensation occurs, and ML predictions are more challenging. Nevertheless, our 2D-EH features still outperform structural features including those derived from persistent homology. The novel 2D-EH features may help accelerate the discovery and design of advanced nanoporous materials using ML for gas storage and separation in the future.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Structure and properties of densified silica glass: characterizing the order within disorder

The broken symmetry in the atomic-scale ordering of glassy versus crystalline solids leads to a daunting challenge to provide suitable metrics for describing the order within disorder, especially on length scales beyond the nearest neighbor that are characterized by rich structural complexity. Here, we address this challenge for silica, a canonical network-forming glass, by using hot versus cold compression to (i) systematically increase the structural ordering after densification and (ii) prepare two glasses with the same high-density but contrasting structures. The structure was measured by high-energy X-ray and neutron diffraction, and atomistic models were generated that reproduce the experimental results. The vibrational and thermodynamic properties of the glasses were probed by using inelastic neutron scattering and calorimetry, respectively. Traditional measures of amorphous structures show relatively subtle changes upon compacting the glass. The method of persistent homology identifies, however, distinct features in the network topology that change as the initially open structure of the glass is collapsed. The results for the same high-density glasses show that the nature of structural disorder does impact the heat capacity and boson peak in the low-frequency dynamical spectra. Densification is discussed in terms of the loss of locally favored tetrahedral structures comprising oxygen-decorated SiSi 4<:sub> tetrahedra.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Cell wall biology of the moss Physcomitrium patens

Abstract The moss Physcomitrium (previously Physcomitrella) patens is a non-vascular plant belonging to the bryophytes that has been used as a model species to study the evolution of plant cell wall structure and biosynthesis. Here, we present an updated review of the cell wall biology of P. patens. Immunocytochemical and structural studies have shown that the cell walls of P. patens mainly contain cellulose, hemicelluloses (xyloglucan, xylan, glucomannan, and arabinoglucan), pectin, and glycoproteins, and their abundance varies among different cell types and at different plant developmental stages. Genetic and biochemical analyses have revealed that a number of genes involved in cell wall biosynthesis are functionally conserved between P. patens and vascular plants, indicating that the common ancestor of mosses and vascular plants had already acquired most of the biosynthetic machinery to make various cell wall polymers. Although P. patens does not synthesize lignin, homologs of the phenylpropanoid biosynthetic pathway genes exist in P. patens and they play an essential role in the production of caffeate derivatives for cuticle formation. Further genetic and biochemical dissection of cell wall biosynthetic genes in P. patens promises to provide additional insights into the evolutionary history of plant cell wall structure and biosynthesis.

Plant Sciences↗
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↗