SEARCH · Engineering Papers
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.
Genomic islands (GIs) are mobile genetic elements that integrate into host genomes via self-encoded integrases at specific DNA sequences known as attachment (att) sites. The ability to predict the target att site from an integrase's protein sequence is a central challenge in genomics due to the sequence diversity of integrases and the subtlety of their DNA recognition motifs.
Proteins are the molecular machines of life with numerous applications in energy, health, and sustainability. However, engineering proteins with desired functions for practical applications remains slow, expensive, and specialist-dependent. Here we report a generally applicable platform for autonomous enzyme engineering that integrates machine learning and large language models with biofoundry automation to eliminate the need for human intervention, judgement, and domain expertise. Requiring only an input protein sequence and a quantifiable way to measure fitness, this automated platform can be applied to engineer a wide array of proteins. As a proof of concept, we engineer Arabidopsis thaliana halide methyltransferase (AtHMT) for a 90-fold improvement in substrate preference and 16-fold improvement in ethyltransferase activity, along with developing a Yersinia mollaretii phytase (YmPhytase) variant with 26-fold improvement in activity at neutral pH. This is accomplished in four rounds over 4 weeks, while requiring construction and characterization of fewer than 500 variants for each enzyme. This platform for autonomous experimentation paves the way for rapid advancements across diverse industries, from medicine and biotechnology to renewable energy and sustainable chemistry.
Proteins are the molecular machines of life with numerous applications in energy, health, and sustainability. However, engineering proteins with desired functions for practical applications remains slow, expensive, and specialist-dependent. Here we report a generally applicable platform for autonomous enzyme engineering that integrates machine learning and large language models with biofoundry automation to eliminate the need for human intervention, judgement, and domain expertise. Requiring only an input protein sequence and a quantifiable way to measure fitness, this automated platform can be applied to engineer a wide array of proteins. As a proof of concept, we engineer Arabidopsis thaliana halide methyltransferase (AtHMT) for a 90-foldimprovement in substrate preference and 16-fold improvement in ethyl-transferase activity, along with developing a Yersinia mollaretii phytase (YmPhytase) variant with 26-fold improvement in activity at neutral pH. This is accomplished in four rounds over 4 weeks, while requiring construction and characterization of fewer than 500 variants for each enzyme. This platform for autonomous experimentation paves the way for rapid advancements across diverse industries, from medicine and biotechnology to renewable energy and sustainable chemistry.
Enzymes are proteins that catalyze specific biochemical reactions in cells. Enzyme Commission (EC) numbers are used to annotate enzymes in a four-level hierarchy that classifies enzymes based on the specific chemical reactions they catalyze. Accurate EC number prediction is essential for understanding enzyme functions. Despite the availability of numerous methods for predicting EC numbers from protein sequences, there is no unified framework for evaluating and studying such methods systematically. This gap limits the ability of the community to identify the most effective approaches for enzyme annotation. We introduce EC-Bench, a benchmark for EC number prediction, consisting of 1) an initial representative set of existing methods (including homology-based, deep learning, contrastive learning, and language model methods), 2) existing and novel accuracy and efficiency performance metrics, and 3) selected datasets to allow for comprehensive comparative study. EC-Bench is open-source and provides a framework for researchers to not only compare among existing methods objectively under uniform conditions, but also to introduce and effectively evaluate performance of new methods in a comparative framework. To demonstrate the utility of EC-Bench, we perform extensive experimentation to compare the existing EC number prediction methods and establish their advantages and disadvantages in a variety of prediction tasks, namely “exact EC number prediction”, “EC number completion” and (partial or additional) “EC number recommendation”. We find wide variation in the performance of different methods, but also subtle but potentially useful differences in the performance of different methods across tasks and for different parts of the EC hierarchy.
Abstract The glycosylation on the spike (S) protein of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that causes COVID-19, modulates the viral infection by altering conformational dynamics, receptor interaction and host immune responses. Several variants of concern (VOCs) of SARS-CoV-2 have evolved during the pandemic, and crucial mutations on the S protein of the virus have led to increased transmissibility and immune escape. In this study, we compare the site-specific glycosylation and overall glycomic profiles of the wild type Wuhan-Hu-1 strain (WT) S protein and five VOCs of SARS-CoV-2: Alpha, Beta, Gamma, Delta and Omicron. Interestingly, both N- and O-glycosylation sites on the S protein are highly conserved among the spike mutant variants, particularly at the sites on the receptor-binding domain (RBD). The conservation of glycosylation sites is noteworthy, as over 2 million SARS-CoV-2 S protein sequences have been reported with various amino acid mutations. Our detailed profiling of the glycosylation at each of the individual sites of the S protein across the variants revealed intriguing possible association of glycosylation pattern on the variants and their previously reported infectivity. While the sites are conserved, we observed changes in the N- and O-glycosylation profile across the variants. The newly emerged variants, which showed higher resistance to neutralizing antibodies and vaccines, displayed a decrease in the overall abundance of complex-type glycans with both fucosylation and sialylation and an increase in the oligomannose-type glycans across the sites. Among the variants, the glycosylation sites with significant changes in glycan profile were observed at both the N -terminal domain and RBD of S protein, with Omicron showing the highest deviation. The increase in oligomannose-type happens sequentially from Alpha through Delta. Interestingly, Omicron does not contain more oligomannose-type glycans compared to Delta but does contain more compared to the WT and other VOCs. O-glycosylation at the RBD showed lower occupancy in the VOCs in comparison to the WT. Our study on the sites and pattern of glycosylation on the SARS-CoV-2 S proteins across the VOCs may help to understand how the virus evolved to trick the host immune system. Our study also highlights how the SARS-CoV-2 virus has conserved both N - and O - glycosylation sites on the S protein of the most successful variants even after undergoing extensive mutations, suggesting a correlation between infectivity/ transmissibility and glycosylation.
Abstract Expressing plant metabolic pathways in microbial platforms is an efficient, cost‐effective solution for producing many desired plant compounds. As eukaryotic organisms, yeasts are often the preferred platform. However, expression of plant enzymes in a yeast frequently leads to failure because the enzymes are poorly adapted to the foreign yeast cellular environment. Here, we first summarize the current engineering approaches for optimizing performance of plant enzymes in yeast. A critical limitation of these approaches is that they are labour‐intensive and must be customized for each individual enzyme, which significantly hinders the establishment of plant pathways in cellular factories. In response to this challenge, we propose the development of a cost‐effective computational pipeline to redesign plant enzymes for better adaptation to the yeast cellular milieu. This proposition is underpinned by compelling evidence that plant and yeast enzymes exhibit distinct sequence features that are generalizable across enzyme families. Consequently, we introduce a data‐driven machine learning framework designed to extract ‘yeastizing’ rules from natural protein sequence variations, which can be broadly applied to all enzymes. Additionally, we discuss the potential to integrate the machine learning model into a full design‐build‐test cycle.
Cellular life relies on enzymes that require metals, which must be acquired from extracellular sources. Bacteria utilize surface and secreted proteins to acquire such valuable nutrients from their environment. These include the cargo proteins of the type eleven secretion system (T11SS), which have been connected to host specificity, metal homeostasis, and nutritional immunity evasion. This Sec-dependent, Gram-negative secretion system is encoded by organisms throughout the phylum Proteobacteria, including human pathogens Neisseria meningitidis, Proteus mirabilis, Acinetobacter baumannii, and Haemophilus influenzae. Experimentally verified T11SS-dependent cargo include transferrin-binding protein B (TbpB), the hemophilin homologs heme receptor protein C (HrpC), hemophilin A (HphA), the immune evasion protein factor-H binding protein (fHbp), and the host symbiosis factor nematode intestinal localization protein C (NilC). Here, we examined the specificity of T11SS systems for their cognate cargo proteins using taxonomically distributed homolog pairs of T11SS and hemophilin cargo and explored the ligand binding ability of those hemophilin cargo homologs. In vivo expression in Escherichia coli of hemophilin homologs revealed that each is secreted in a specific manner by its cognate T11SS protein. Sequence analysis and structural modeling suggest that all hemophilin homologs share an N-terminal ligand-binding domain with the same topology as the ligand-binding domains of the Haemophilus haemolyticus heme binding protein (Hpl) and HphA. We term this signature feature of this group of proteins the hemophilin ligand-binding domain. Network analysis of hemophilin homologs revealed five subclusters and representatives from four of these showed variable heme-binding activities, which, combined with sequence-structure variation, suggests that hemophilins are diversifying in function.
Abstract Background Estimation of the accuracy (quality) of protein structural models is important for both prediction and use of protein structural models. Deep learning methods have been used to integrate protein structure features to predict the quality of protein models. Inter-residue distances are key information for predicting protein’s tertiary structures and therefore have good potentials to predict the quality of protein structural models. However, few methods have been developed to fully take advantage of predicted inter-residue distance maps to estimate the accuracy of a single protein structural model. Result We developed an attentive 2D convolutional neural network (CNN) with channel-wise attention to take only a raw difference map between the inter-residue distance map calculated from a single protein model and the distance map predicted from the protein sequence as input to predict the quality of the model. The network comprises multiple convolutional layers, batch normalization layers, dense layers, and Squeeze-and-Excitation blocks with attention to automatically extract features relevant to protein model quality from the raw input without using any expert-curated features. We evaluated DISTEMA’s capability of selecting the best models for CASP13 targets in terms of ranking loss of GDT-TS score. The ranking loss of DISTEMA is 0.079, lower than several state-of-the-art single-model quality assessment methods. Conclusion This work demonstrates that using raw inter-residue distance information with deep learning can predict the quality of protein structural models reasonably well. DISTEMA is freely at https://github.com/jianlin-cheng/DISTEMA
Conflicting biological goals often meet in the specification of protein sequences for structure and function. Overall, strong energetic conflicts are minimized in folded native states according to the principle of minimal frustration, so that a sequence can spontaneously fold, but local violations of this principle open up the possibility to encode the complex energy landscapes that are required for active biological functions. We survey the local energetic frustration patterns of all protein enzymes with known structures and experimentally annotated catalytic residues. In agreement with previous hypotheses, the catalytic sites themselves are often highly frustrated regardless of the protein oligomeric state, overall topology, and enzymatic class. At the same time a secondary shell of more weakly frustrated interactions surrounds the catalytic site itself. We evaluate the conservation of these energetic signatures in various family members of major enzyme classes, showing that local frustration is evolutionarily more conserved than the primary structure itself.
Algae are a broad class of photosynthetic eukaryotes that are phylogenetically and physiologically diverse. Most of the phylogenetic diversity has been inferred from 18S rDNA sequencing since there are only a few complete genomes available in public databases. Here we use ultra-long-read Nanopore sequencing to determine a gapless, telomere-to-telomere complete genome sequence of Chlorella sp. SLA-04, previously described as Chlorella sorokiniana SLA-04. Chlorella sp. SLA-04 is a green alga that grows to high cell density in a wide variety of environments - high and neutral pH, high and low alkalinity, and high and low salinity. SLA-04's ability to grow in high pH and high alkalinity media without external CO 2 supply is favorable for large-scale algal biomass production. Phylogenetic analysis performed using ribosomal DNA and conserved protein sequences consistently reveal that Chlorella sp. SLA-04 forms a distinct lineage from other strains of Chlorella sorokiniana. We complement traditional genome annotation methods with high throughput structural predictions and demonstrate that this approach expands functional prediction of the SLA-04 proteome. Genomic analysis of the SLA-04 genome identifies the genes capable of utilizing TCA cycle intermediates to replenish cytosolic acetyl-CoA pools for lipid production. We also identify a complete metabolic pathway for sphingolipid anabolism that may allow SLA-04 to readily adapt to changing environmental conditions and facilitate robust cultivation in mass production systems. Altogether, this work clarifies the phylogeny of Chlorella sp. SLA-04 within Trebouxiophyceae and demonstrates how structural predictions can be used to improve annotation beyond sequencebased methods.
This dataset contains the structural models for the primary transcripts of the Sphagnum divinum proteome. Additionally, for a subset of these proteins, sequence and structural alignment results are provided. This dataset represents the most thorough structural study of a Sphagnum species, also known as peat mosses, by providing three-dimensional atomic resolution structures of the majority of the encoded proteins as well as structural alignment results used in the application of annotating the proteome. References (DOI) AlphaFold v2 Monomer: https://doi.org/10.1038/s41586-021-03819-2. References (DOI) US-align2: https://doi.org/10.1038/s41592-022-01585-1
Top-down mass spectrometry (TD-MS) generates fragment ions that returns information on the polypeptide amino acid sequence. In addition to terminal fragments, internal fragments that result from multiple cleavage events can also be formed. Traditionally, internal fragments are largely ignored due to a lack of available software to reliably assign them, mainly caused by a poor understanding of their formation mechanism. To accurately assign internal fragments, their formation process needs to be better understood. Here we applied a statistical method to compare fragmentation patterns of internal and terminal fragments of peptides and proteins generated by collisionally activated dissociation (CAD). Internal fragments share similar fragmentation propensities with terminal fragments (e.g., enhanced cleavages N-terminal to proline and C-terminal to acidic residues), suggesting that their formation follows conventional CAD pathways. Internal fragments should be generated by subsequent cleavages of terminal fragments and their formation can be explained by the well-known mobile proton model. Additionally, internal fragments can be coupled with terminal fragments to form complementary product ions that span the entire protein sequence. These enhance our understanding of internal fragment formation and can help improve sequencing algorithms to accurately assign internal fragments, which will ultimately lead to more efficient and comprehensive TD-MS analysis of proteins and proteoforms.
The mutation of SARS-CoV-2 influences viral function as residue replacements affect both physiochemical properties and folding conformations. Although a large amount of data on SARS-CoV-2 is available, the investigation of how viral functions change in response to mutations is hampered by a lack of effective structural analysis. Here, we exploit the advances of protein structure fingerprint technology to study the folding conformational changes induced by mutations. With integration of both protein sequences and folding conformations, the structures are aligned for SARS-CoV to SARS-CoV-2, including Alpha variant (lineage B.1.1.7) and Delta variant (lineage B.1.617.2). The results showed that the virus evolution with change in mutational positions and physicochemical properties increased the affinity between spike protein and ACE2, which plays a critical role in coronavirus entry into human cells. Additionally, these structural variations impact vaccine effectiveness and drug function over the course of SARS-CoV-2 evolution. The analysis of structural variations revealed how the coronavirus has gradually evolved in both structure and function and how the SARS-CoV-2 variants have contributed to more severe acute disease worldwide.
Abstract Protein–ligand interactions (PLIs) are essential for biochemical functionality and their identification is crucial for estimating biophysical properties for rational therapeutic design. Currently, experimental characterization of these properties is the most accurate method, however, this is very time-consuming and labor-intensive. A number of computational methods have been developed in this context but most of the existing PLI prediction heavily depends on 2D protein sequence data. Here, we present a novel parallel graph neural network (GNN) to integrate knowledge representation and reasoning for PLI prediction to perform deep learning guided by expert knowledge and informed by 3D structural data. We develop two distinct GNN architectures: $$\hbox {GNN}_{\mathrm{F}}$$ GNN F is the base implementation that employs distinct featurization to enhance domain-awareness, while $$\hbox {GNN}_{\mathrm{P}}$$ GNN P is a novel implementation that can predict with no prior knowledge of the intermolecular interactions. The comprehensive evaluation demonstrated that GNN can successfully capture the binary interactions between ligand and protein’s 3D structure with 0.979 test accuracy for $$\hbox {GNN}_{\mathrm{F}}$$ GNN F and 0.958 for $$\hbox {GNN}_{\mathrm{P}}$$ GNN P for predicting activity of a protein–ligand complex. These models are further adapted for regression tasks to predict experimental binding affinities and $$\hbox {pIC}_{\mathrm{50}}$$ pIC 50 crucial for compound’s potency and efficacy. We achieve a Pearson correlation coefficient of 0.66 and 0.65 on experimental affinity and 0.50 and 0.51 on $$\hbox {pIC}_{\mathrm{50}}$$ pIC 50 with $$\hbox {GNN}_{\mathrm{F}}$$ GNN F and $$\hbox {GNN}_{\mathrm{P}}$$ GNN P , respectively, outperforming similar 2D sequence based models. Our method can serve as an interpretable and explainable artificial intelligence (AI) tool for predicted activity, potency, and biophysical properties of lead candidates. To this end, we show the utility of $$\hbox {GNN}_{\mathrm{P}}$$ GNN P on SARS-Cov-2 protein targets by screening a large compound library and comparing the prediction with the experimentally measured data.
Summary Wildfires drastically impact the soil environment, altering the soil organic matter, forming pyrolyzed compounds, and markedly reducing the diversity of microorganisms. Pyrophilous fungi, especially the species from the orders Pezizales and Agaricales, are fire‐responsive fungal colonizers of post‐fire soil that have historically been found fruiting on burned soil and thus may encode mechanisms of processing these compounds in their genomes. Pyrophilous fungi are diverse. In this work, we explored this diversity and sequenced six new genomes of pyrophilous Pezizales fungi isolated after the 2013 Rim Fire near Yosemite Park in California, USA: Pyronema domesticum , Pyronema omphalodes , Tricharina praecox , Geopyxis carbonaria , Morchella snyderi , and Peziza echinospora . A comparative genomics analysis revealed the enrichment of gene families involved in responses to stress and the degradation of pyrolyzed organic matter. In addition, we found that both protein sequence lengths and G + C content in the third base of codons (GC3) in pyrophilous fungi fall between those in mesophilic/nonpyrophilous and thermophilic fungi. A comparative transcriptome analysis of P. domesticum under two conditions – growing on charcoal, and during sexual development – identified modules of genes that are co‐expressed in the charcoal and light‐induced sexual development conditions. In addition, environmental sensors such as transcription factors STE12, LreA, LreB, VosA, and EsdC were upregulated in the charcoal condition. Taken together, these results highlight genomic adaptations of pyrophilous fungi and indicate a potential connection between charcoal tolerance and fruiting body formation in P. domesticum .
Abstract Noncanonical cofactor biomimetics (NCBs) such as nicotinamide mononucleotide (NMN + ) provide enhanced scalability for biomanufacturing. However, engineering enzymes to accept NCBs is difficult. Here, we establish a growth selection platform to evolve enzymes to utilize NMN + -based reducing power. This is based on an orthogonal, NMN + -dependent glycolytic pathway in Escherichia coli which can be coupled to any reciprocal enzyme to recycle the ensuing reduced NMN + . With a throughput of >10 6 variants per iteration, the growth selection discovers a Lactobacillus pentosus NADH oxidase variant with ~10-fold increase in NMNH catalytic efficiency and enhanced activity for other NCBs. Molecular modeling and experimental validation suggest that instead of directly contacting NCBs, the mutations optimize the enzyme’s global conformational dynamics to resemble the WT with the native cofactor bound. Restoring the enzyme’s access to catalytically competent conformation states via deep navigation of protein sequence space with high-throughput evolution provides a universal route to engineer NCB-dependent enzymes.
BACKGROUND: The pattern and timing of the rise in complex multicellular life during Earth's history has not been established. Great disparity persists between the pattern suggested by the fossil record and that estimated by molecular clocks, especially for plants, animals, fungi, and the deepest branches of the eukaryote tree. Here, we used all available protein sequence data and molecular clock methods to place constraints on the increase in complexity through time. RESULTS: Our phylogenetic analyses revealed that (i) animals are more closely related to fungi than to plants, (ii) red algae are closer to plants than to animals or fungi, (iii) choanoflagellates are closer to animals than to fungi or plants, (iv) diplomonads, euglenozoans, and alveolates each are basal to plants+animals+fungi, and (v) diplomonads are basal to other eukaryotes (including alveolates and euglenozoans). Divergence times were estimated from global and local clock methods using 20-188 proteins per node, with data treated separately (multigene) and concatenated (supergene). Different time estimation methods yielded similar results (within 5%): vertebrate-arthropod (964 million years ago, Ma), Cnidaria-Bilateria (1,298 Ma), Porifera-Eumetozoa (1,351 Ma), Pyrenomycetes-Plectomycetes (551 Ma), Candida-Saccharomyces (723 Ma), Hemiascomycetes-filamentous Ascomycota (982 Ma), Basidiomycota-Ascomycota (968 Ma), Mucorales-Basidiomycota (947 Ma), Fungi-Animalia (1,513 Ma), mosses-vascular plants (707 Ma), Chlorophyta-Tracheophyta (968 Ma), Rhodophyta-Chlorophyta+Embryophyta (1,428 Ma), Plantae-Animalia (1,609 Ma), Alveolata-plants+animals+fungi (1,973 Ma), Euglenozoa-plants+animals+fungi (1,961 Ma), and Giardia-plants+animals+fungi (2,309 Ma). By extrapolation, mitochondria arose approximately 2300-1800 Ma and plastids arose 1600-1500 Ma. Estimates of the maximum number of cell types of common ancestors, combined with divergence times, showed an increase from two cell types at 2500 Ma to approximately 10 types at 1500 Ma and 50 cell types at approximately 1000 Ma. CONCLUSIONS: The results suggest that oxygen levels in the environment, and the ability of eukaryotes to extract energy from oxygen, as well as produce oxygen, were key factors in the rise of complex multicellular life. Mitochondria and organisms with more than 2-3 cell types appeared soon after the initial increase in oxygen levels at 2300 Ma. The addition of plastids at 1500 Ma, allowing eukaryotes to produce oxygen, preceded the major rise in complexity.
SPARC facilitates the generation of plausible hypotheses regarding underlying biochemical mechanisms by structurally characterizing protein sequence constraints. Such constraints appear as residues co-conserved in functionally related subgroups, as subtle pairwise correlations (i.e., direct couplings), and as correlations among these sequence features or with structural features. SPARC performs three types of analyses. First, based on pairwise sequence correlations, it estimates the biological relevance of alternative conformations and of homomeric contacts, as illustrated here for death domains. Second, it estimates the statistical significance of the correspondence between directly coupled residue pairs and interactions at heterodimeric interfaces. Third, given molecular dynamics simulated structures, it characterizes interactions among constrained residues or between such residues and ligands that: (a) are stably maintained during the simulation; (b) undergo correlated formation and/or disruption of interactions with other constrained residues; or (c) switch between alternative interactions. We illustrate this for two homohexameric complexes: the bacterial enhancer binding protein (bEBP) NtrC1, which activates transcription by remodeling RNA polymerase (RNAP) containing σ 54 , and for DnaB helicase, which opens DNA at the bacterial replication fork. Based on the NtrC1 analysis, we hypothesize possible mechanisms for inhibiting ATP hydrolysis until ADP is released from an adjacent subunit and for coupling ATP hydrolysis to restructuring of σ 54 binding loops. Based on the DnaB analysis, we hypothesize that DnaB ‘grabs’ ssDNA by flipping every fourth base and inserting it into cavities between subunits and that flipping of a DnaB-specific glutamine residue triggers ATP hydrolysis.