Homologous Pairs of Low and High Temperature Originating Proteins Spanning the Known Prokaryotic Universe.

Stability of proteins at high temperature has been a topic of interest for many years, as this attribute is favourable for applications ranging from therapeutics to industrial chemical manufacturing. Our current understanding and methods for designing high-temperature stability into target proteins are inadequate. To drive innovation in this space, we have curated a large dataset, learn2thermDB, of protein-temperature examples, totalling 24 million instances, and paired proteins across temperatures based on homology, yielding 69 million protein pairs - orders of magnitude larger than the current largest. This important step of pairing allows for study of high-temperature stability in a sequence-dependent manner in the big data era. The data pipeline is parameterized and open, allowing it to be tuned by downstream users. We further show that the data contains signal for deep learning. This data offers a new doorway towards thermal stability design models.

Efficient 3D Molecular Design with an E(3) Invariant Transformer VAE.

This work introduces a three-dimensional (3D) invariant graph-to-string transformer variational autoencoders (VAE) (Vagrant) for generating molecules with accurate density functional theory (DFT)-level properties. Vagrant learns to model the joint probability distribution of a 3D molecular structure and its properties by encoding molecular structures into a 3D-aware latent space. Directed navigation through this latent space implicitly optimizes the 3D structure of a molecule, and the latent embedding can be used to condition a generative transformer to predict the candidate structure as a one-dimensional (1D) sequence. Additionally, we introduce two novel sampling methods that exploit the latent characteristics of a VAE to improve performance. We show that our method outperforms comparable 3D autoregressive and diffusion methods for predicting quantum chemical property values of novel molecules in terms of both sample quality and computational efficiency.

Mass-Suite: a novel open-source python package for high-resolution mass spectrometry data analysis.

Mass-Suite (MSS) is a Python-based, open-source software package designed to analyze high-resolution mass spectrometry (HRMS)-based non-targeted analysis (NTA) data, particularly for water quality assessment and other environmental applications. MSS provides flexible, user-defined workflows for HRMS data processing and analysis, including both basic functions (e.g., feature extraction, data reduction, feature annotation, data visualization, and statistical analyses) and advanced exploratory data mining and predictive modeling capabilities that are not provided by currently available open-source software (e.g., unsupervised clustering analyses, a machine learning-based source tracking and apportionment tool). As a key advance, most core MSS functions are supported by machine learning algorithms (e.g., clustering algorithms and predictive modeling algorithms) to facilitate function accuracy and/or efficiency. MSS reliability was validated with mixed chemical standards of known composition, with 99.5% feature extraction accuracy and ~ 52% overlap of extracted features relative to other open-source software tools. Example user cases of laboratory data evaluation are provided to illustrate MSS functionalities and demonstrate reliability. MSS expands available HRMS data analysis workflows for water quality evaluation and environmental forensics, and is readily integrated with existing capabilities. As an open-source package, we anticipate further development of improved data analysis capabilities in collaboration with interested users.

A methanotrophic bacterium to enable methane removal for climate mitigation.

The rapid increase of the potent greenhouse gas methane in the atmosphere creates great urgency to develop and deploy technologies for methane mitigation. One approach to removing methane is to use bacteria for which methane is their carbon and energy source (methanotrophs). Such bacteria naturally convert methane to CO2 and biomass, a value-added product and a cobenefit of methane removal. Typically, methanotrophs grow best at around 5,000 to 10,000 ppm methane, but methane in the atmosphere is 1.9 ppm. Air above emission sites such as landfills, anaerobic digestor effluents, rice paddy effluents, and oil and gas wells contains elevated methane in the 500 ppm range. If such sites are targeted for methane removal, technology harnessing aerobic methanotroph metabolism has the potential to become economically and environmentally viable. The first step in developing such methane removal technology is to identify methanotrophs with enhanced ability to grow and consume methane at 500 ppm and lower. We report here that some existing methanotrophic strains grow well at 500 ppm methane, and one of them, Methylotuvimicrobium buryatense 5GB1C, consumes such low methane at enhanced rates compared to previously published values. Analyses of bioreactor-based performance and RNAseq-based transcriptomics suggest that this ability to utilize low methane is based at least in part on extremely low non-growth-associated maintenance energy and on high methane specific affinity. This bacterium is a candidate to develop technology for methane removal at emission sites. If appropriately scaled, such technology has the potential to slow global warming by 2050.

Quick and Accurate Estimates of Mutation Effects on Transition-State Stabilization of Enzymes from Molecular Simulations with Restrained Transition States.

Data science and machine learning are revolutionizing enzyme engineering; however, high-throughput simulations for screening large libraries of enzyme variants remain a challenge. Here, we present a novel but highly simple approach to comparing enzyme variants with fully atomistic classical molecular dynamics (MD) simulations on a tractable timescale. Our method greatly simplifies the problem by restricting sampling only to the reaction transition state, and we show that the resulting measurements of transition-state stability are well correlated with experimental activity measurements across two highly distinct enzymes, even for mutations with effects too small to resolve with free energy methods. This method will enable atomistic simulations to achieve sampling coverage for enzyme variant prescreening and machine learning model training on a scale that was previously not possible.

Water-Accelerated Photooxidation of CH3NH3PbI3 Perovskite.

Halide perovskites have the potential to disrupt the photovoltaics market based on their high performance and low cost. However, the decomposition of perovskites under moisture, oxygen, and light raises concerns about service lifetime, especially because degradation mechanisms and the corresponding rate laws that fit the observed data have thus far eluded researchers. Here, we report a water-accelerated photooxidation mechanism dominating the degradation kinetics of archetypal perovskite CH3NH3PbI3 in air under >1% relative humidity at 25 °C. From this mechanism, we develop a kinetic model that quantitatively predicts the degradation rate as a function of temperature, ambient O2 and H2O levels, and illumination. Because water is a possible product of dry photooxidation, these results highlight the need for encapsulation schemes that rigorously block oxygen ingress, as product water may accumulate beneath the encapsulant and initiate the more rapid water-accelerated photooxidative decomposition.

An automated multiplexed turbidometric and data collection system for measuring growth kinetics of anaerobes dependent on gaseous substrates.

Standard methods of monitoring the growth kinetics of anaerobic microorganisms are generally impractical when there is a protracted or indeterminate period of active growth, and when high numbers of samples or replications are required. As part of our studies of the adaptive evolution of a simple anaerobic syntrophic mutualism, requiring the characterization of many isolates and alternative syntrophic pairings, we developed a multiplexed growth monitoring system using a combination of commercially available electronics and custom designed circuitry and materials. This system automatically monitors up to 64 sealed, and as needed pressurized, culture tubes and reports the growth data in real-time through integration with a customized relational database. The utility of this system was demonstrated by resolving minor differences in growth kinetics associated with the adaptive evolution of a simple microbial community comprised of a sulfate reducing bacterium, Desulfovibrio vulgaris, grown in syntrophic association with Methanococcus maripaludis, a hydrogenotrophic methanogen.

Attention-based generative models for de novo molecular design.

Attention mechanisms have led to many breakthroughs in sequential data modeling but have yet to be incorporated into any generative algorithms for molecular design. Here we explore the impact of adding self-attention layers to generative ß-VAE models and show that those with attention are able to learn a complex "molecular grammar" while improving performance on downstream tasks such as accurately sampling from the latent space ("model memory") or exploring novel chemistries not present in the training data. There is a notable relationship between a model's architecture, the structure of its latent memory and its performance during inference. We demonstrate that there is an unavoidable tradeoff between model exploration and validity that is a function of the complexity of the latent memory. However, novel sampling schemes may be used that optimize this tradeoff. We anticipate that attention will play an important role in future molecular design algorithms that can make efficient use of the detailed molecular substructures learned by the transformer.

A Computational Framework for Identifying Promoter Sequences in Nonmodel Organisms Using RNA-seq Data Sets.

Engineering microorganisms into biological factories that convert renewable feedstocks into valuable materials is a major goal of synthetic biology; however, for many nonmodel organisms, we do not yet have the genetic tools, such as suites of strong promoters, necessary to effectively engineer them. In this work, we developed a computational framework that can leverage standard RNA-seq data sets to identify sets of constitutive, strongly expressed genes and predict strong promoter signals within their upstream regions. The framework was applied to a diverse collection of RNA-seq data measured for the methanotroph Methylotuvimicrobium buryatense 5GB1 and identified 25 genes that were constitutively, strongly expressed across 12 experimental conditions. For each gene, the framework predicted short (27-30 nucleotide) sequences as candidate promoters and derived -35 and -10 consensus promoter motifs (TTGACA and TATAAT, respectively) for strong expression in M. buryatense. This consensus closely matches the canonical E. coli sigma-70 motif and was found to be enriched in promoter regions of the genome. A subset of promoter predictions was experimentally validated in a XylE reporter assay, including the consensus promoter, which showed high expression. The pmoC, pqqA, and ssrA promoter predictions were additionally screened in an experiment that scrambled the -35 and -10 signal sequences, confirming that transcription initiation was disrupted when these specific regions of the predicted sequence were altered. These results indicate that the computational framework can make biologically meaningful promoter predictions and identify key pieces of regulatory systems that can serve as foundational tools for engineering diverse microorganisms for biomolecule production.

Data Science in Chemical Engineering: Applications to Molecular Science.

Chemical engineering is being rapidly transformed by the tools of data science. On the horizon, artificial intelligence (AI) applications will impact a huge swath of our work, ranging from the discovery and design of new molecules to operations and manufacturing and many areas in between. Early adoption of data science, machine learning, and early examples of AI in chemical engineering has been rich with examples of molecular data science-the application tools for molecular discovery and property optimization at the atomic scale. We summarize key advances in this nascent subfield while introducing molecular data science for a broad chemical engineering readership. We introduce the field through the concept of a molecular data science life cycle and discuss relevant aspects of five distinct phases of this process: creation of curated data sets, molecular representations, data-driven property prediction, generation of new molecules, and feasibility and synthesizability considerations.

Continuous Molecular Representations of Ionic Liquids.

Designing new ionic liquids (ILs) is of crucial importance for various industrial applications. However, this always leads to a daunting challenge, as the number of possible combinations of cation and anion are very high and it is impossible to experimentally propose and screen a wide pool of potential candidates. However, recent applications of machine learning (ML) models have greatly improved the overall chemical discovery pipeline. In this study, we compare different generative methods for producing ionic liquids. In this comparison, we show the following: (1) when training data is scarce, a transfer learning approach can be applied to variational autoencoders (VAEs) to generate molecular structures of the target molecule type; (2) in a VAE-like structure, separate latent spaces for the cationic and anionic moieties can result in meaningful representations for their combinative, macroscopic properties; (3) interpolating between ILs with desired properties can result in a new IL with attributes similar to the two structural end points.

Core Metabolism Shifts during Growth on Methanol versus Methane in the Methanotroph Methylomicrobium buryatense 5GB1.

Methylomicrobium buryatense 5GB1 is an obligate methylotroph which grows on methane or methanol with similar growth rates. It has long been assumed that the core metabolic pathways must be similar on the two substrates, but recent studies of methane metabolism in this bacterium suggest that growth on methanol might have significant differences from growth on methane. In this study, both a targeted metabolomics approach and a 13C tracer approach were taken to understand core carbon metabolism in M. buryatense 5GB1 during growth on methanol and to determine whether such differences occur. Our results suggest a systematic shift of active core metabolism in which increased flux occurred through both the Entner-Doudoroff (ED) pathway and the partial serine cycle, while the tricarboxylic acid (TCA) cycle was incomplete, in contrast to growth on methane. Using the experimental results as constraints, we applied flux balance analysis to determine the metabolic flux phenotype of M. buryatense 5GB1 growing on methanol, and the results are consistent with predictions based on ATP and NADH changes. Transcriptomics analysis suggested that the changes in fluxes and metabolite levels represented results of posttranscriptional regulation. The combination of flux balance analysis of the genome-scale model and the flux ratio from 13C data changed the solution space for a better prediction of cell behavior and demonstrated the significant differences in physiology between growth on methane and growth on methanol.IMPORTANCE One-carbon compounds such as methane and methanol are of increasing interest as sustainable substrates for biological production of fuels and industrial chemicals. The bacteria that carry out these conversions have been studied for many decades, but gaps exist in our knowledge of their metabolic pathways. One such gap is the difference between growth on methane and growth on methanol. Understanding such metabolism is important, since each has advantages and disadvantages as a feedstock for production of chemicals and fuels. The significance of our research is in the demonstration that the metabolic network is substantially altered in each case and in the delineation of these changes. The resulting new insights into the core metabolism of this bacterium now provide an improved basis for future strain design.

Mosaic plasmids are abundant and unevenly distributed across prokaryotic taxa.

Mosaic plasmids, plasmids composed of genetic elements from distinct sources, are associated with the spread of antibiotic resistance genes. Transposons are considered the primary mechanism for mosaic plasmid formation, though other mechanisms have been observed in specific instances. The frequency with which mosaic plasmids have been described suggests they may play an important role in plasmid population dynamics. Our survey of the confirmed plasmid sequences available from complete and draft genomes in the RefSeq database shows that 46% of them fit a strict definition of mosaic. Mosaic plasmids are also not evenly distributed over the taxa represented in the database. Plasmids from some genera, including Piscirickettsia and Yersinia, are almost all mosaic, while plasmids from other genera, including Borrelia, are rarely mosaic. While some mosaic plasmids share identical regions with hundreds of others, the median mosaic plasmid only shares with 8 other plasmids. When considering only plasmids from finished genomes (51.6% of the total), mosaic plasmids have significantly higher proportions of transposase and antibiotic resistance genes. Conversely, only 56.6% of mosaic fragments (DNA fragments shared between mosaic plasmids) contain a recognizable transposase gene, and only 1.2% of mosaic fragments are flanked by inverted repeats. Mosaic fragments associated with the IS26 transposase gene are 3.8-fold more abundant than any other sequence shared between mosaic plasmids in the database, though this is at least partly due to overrepresentation of Enterobacteriaceae plasmids. Mosaic plasmids are a complicated trait of some plasmid populations, only partly explained by transposition. Though antibiotic resistance genes led to the identification of many mosaic plasmids, mosaic plasmids are a broad phenomenon encompassing many more traits than just antibiotic resistance. Further research will be required to determine the influence of ecology, host repair mechanisms, conjugation, and plasmid host range on the formation and influence of mosaic plasmids. AUTHOR SUMMARY: Plasmids are extrachromosomal genetic entities that are found in many prokaryotes. They serve as flexible storage for genes, and individual cells can make substantial changes to their characteristics by acquiring, losing, or modifying a plasmid. In some pathogenic bacteria, such as Escherichia coli, antibiotic resistance genes are known to spread primarily on plasmids. By analyzing a database of 8592 plasmid sequences we determined that many of these plasmids have exchanged genes with each other, becoming mosaics of genes from different sources. We next separated these plasmids into groups based on the organism they were isolated from and found that different groups had different fractions of mosaic plasmids. This result was unexpected and suggests that the mechanisms and selective pressures causing mosaic plasmids do not occur evenly over all species. It also suggests that plasmids may provide different levels of potential variation to different species. This work uncovers a previously unrecognized pattern in plasmids across prokaryotes, that could lead to new insights into the evolutionary role that plasmids play.

DNA-SIP based genome-centric metagenomics identifies key long-chain fatty acid-degrading populations in anaerobic digesters with different feeding frequencies.

Fats, oils and greases (FOG) are energy-dense wastes that can be added to anaerobic digesters to substantially increase biomethane recovery via their conversion through long-chain fatty acids (LCFAs). However, a better understanding of the ecophysiology of syntrophic LCFA-degrading microbial communities in anaerobic digesters is needed to develop operating strategies that mitigate inhibitory LCFA accumulation from FOG. In this research, DNA stable isotope probing (SIP) was coupled with metagenomic sequencing for a genome-centric comparison of oleate (C18:1)-degrading populations in two anaerobic codigesters operated with either a pulse feeding or continuous-feeding strategy. The pulse-fed codigester microcosms converted oleate into methane at over 20% higher rates than the continuous-fed codigester microcosms. Differential coverage binning was demonstrated for the first time to recover population genome bins (GBs) from DNA-SIP metagenomes. About 70% of the 13C-enriched GBs were taxonomically assigned to the Syntrophomonas genus, thus substantiating the importance of Syntrophomonas species to LCFA degradation in anaerobic digesters. Phylogenetic comparisons of 13C-enriched GBs showed that phylogenetically distinct Syntrophomonas GBs were unique to each codigester. Overall, these results suggest that syntrophic populations in anaerobic digesters can have different adaptive capacities, and that selection for divergent populations may be achieved by adjusting reactor operating conditions to maximize biomethane recovery.

Natural Selection in Synthetic Communities Highlights the Roles of Methylococcaceae and Methylophilaceae and Suggests Differential Roles for Alternative Methanol Dehydrogenases in Methane Consumption.

We describe experiments that follow species dynamics and gene expression patterns in synthetic bacterial communities including species that compete for the single carbon substrate supplied, methane, and species unable to consume methane, which could only succeed through cooperative interactions. We demonstrate that these communities mostly select for two functional guilds, methanotrophs of the family Methylococcaceae and non-methanotrophic methylotrophs of the family Methylophilaceae, these taxonomic guilds outcompeting all other species included in the synthetic mix. The metatranscriptomics analysis uncovered that in both Methylococcaceae and Methylophilaceae, some of the most highly transcribed genes were the ones encoding methanol dehydrogenases (MDH). Remarkably, expression of alternative MDH genes (mxaFI versus xoxF), previously shown to be subjects to the rare Earth element switch, was found to depend on environmental conditions such as nitrogen source and methane and O2 partial pressures, and also to be species-specific. Along with the xoxF genes, genes encoding divergent cytochromes were highly expressed in both Methylophilaceae and Methylococcaceae, suggesting their function in methanol metabolism, likely encoding proteins serving as electron acceptors from XoxF enzymes. The research presented tested a synthetic community model that is much simplified compared to natural communities consuming methane, but more complex than the previously utilized two-species model. The performance of this model identifies prominent species for future synthetic ecology experiments and highlights both advantages of this approach and the challenges that it presents.

Oxygen-limited metabolism in the methanotroph Methylomicrobium buryatense 5GB1C.

The bacteria that grow on methane aerobically (methanotrophs) support populations of non-methanotrophs in the natural environment by excreting methane-derived carbon. One group of excreted compounds are short-chain organic acids, generated in highest abundance when cultures are grown under O2-starvation. We examined this O2-starvation condition in the methanotroph Methylomicrobium buryatense 5GB1. The M. buryatense 5GB1 genome contains homologs for all enzymes necessary for a fermentative metabolism, and we hypothesize that a metabolic switch to fermentation can be induced by low-O2 conditions. Under prolonged O2-starvation in a closed vial, this methanotroph increases the amount of acetate excreted about 10-fold, but the formate, lactate, and succinate excreted do not respond to this culture condition. In bioreactor cultures, the amount of each excreted product is similar across a range of growth rates and limiting substrates, including O2-limitation. A set of mutants were generated in genes predicted to be involved in generating or regulating excretion of these compounds and tested for growth defects, and changes in excretion products. The phenotypes and associated metabolic flux modeling suggested that in M. buryatense 5GB1, formate and acetate are excreted in response to redox imbalance. Our results indicate that even under O2-starvation conditions, M. buryatense 5GB1 maintains a metabolic state representing a combination of fermentation and respiration metabolism.

Metabolic crosstalk regulates Porphyromonas gingivalis colonization and virulence during oral polymicrobial infection.

Many human infections are polymicrobial in origin, and interactions among community inhabitants shape colonization patterns and pathogenic potential 1 . Periodontitis, which is the sixth most prevalent infectious disease worldwide 2 , ensues from the action of dysbiotic polymicrobial communities 3 . The keystone pathogen Porphyromonas gingivalis and the accessory pathogen Streptococcus gordonii interact to form communities in vitro and exhibit increased fitness in vivo 3,4 . The mechanistic basis of this polymicrobial synergy, however, has not been fully elucidated. Here we show that streptococcal 4-aminobenzoate/para-amino benzoic acid (pABA) is required for maximal accumulation of P. gingivalis in dual-species communities. Metabolomic and proteomic data showed that exogenous pABA is used for folate biosynthesis, and leads to decreased stress and elevated expression of fimbrial adhesins. Moreover, pABA increased the colonization and survival of P. gingivalis in a murine oral infection model. However, pABA also caused a reduction in virulence in vivo and suppressed extracellular polysaccharide production by P. gingivalis. Collectively, these data reveal a multidimensional aspect to P. gingivalis-S. gordonii interactions and establish pABA as a critical cue produced by a partner species that enhances the fitness of P. gingivalis while diminishing its virulence.

Long-chain fatty acid feeding frequency in anaerobic codigestion impacts syntrophic community structure and biokinetics.

This study investigated the impacts of long-chain fatty acid (LCFA) feeding frequencies on microbial community structure, bioconversion kinetics, and process stability during anaerobic codigestion. Parallel laboratory-scale anaerobic codigesters fed with dairy cattle manure were either pulse-fed every two days or continuously-fed daily, respectively, with oleate (C18:1) in incremental step increases over 200 days up to 64% of the influent chemical oxygen demand (COD). The effluent acetate concentration exceeded 3000 mg/L in the continuous-fed codigester at the highest oleate loading rate, but remained below 100 mg/L in the pulse-fed codigester at the end of its 48-hr oleate feed cycle. Maximum substrate conversion rates of oleate (qmax, oleate) and acetate (qmax, acetate) were significantly higher in the pulse-fed codigester compared to the continuous-fed codigester. 16S rRNA gene amplicon sequencing showed that Bacteria and Archaea community profiles diverged based on the codigester LCFA feeding pattern and loading rate. LCFA-degrading Syntrophomonas bacteria were significantly enriched in both LCFA codigesters relative to the control digester. The pulse-fed codigester had the highest community fraction of Syntrophomonas 16S rRNA genes by the end of the experiment with 43% of Bacteria amplicon sequences. qmax, oleate and qmax, acetate values were both significantly correlated to absolute concentrations of Syntrophomonas and Methanosaeta 16S rRNA genes, respectively. Multiple-linear regression models based on the absolute abundance of Syntrophomonas and Methanosaeta taxa provided improved predictions of oleate and acetate bioconversion kinetics, respectively. These results collectively suggest that pulse feeding rather than continuous feeding LCFA during anaerobic codigestion selected for higher microbial bioconversion kinetics and functional stability, which were related to changes in the physiological diversity and adaptive capacity of syntrophic and methanogenic communities.

Insights into Dynamic Polymicrobial Synergy Revealed by Time-Coursed RNA-Seq.

Many bacterial infections involve polymicrobial communities in which constituent organisms are synergistically pathogenic. Periodontitis, a commonly occurring chronic inflammatory disorder, is induced by multispecies bacterial communities. The periodontal keystone pathogen Porphyromonas gingivalis and the accessory pathogen Streptococcus gordonii exhibit polymicrobial synergy in animal models of disease. Mechanisms of co-adhesion and community formation by P. gingivalis and S. gordonii are well-established; however, little is known regarding the basis for increased pathogenicity. In this study we used time-coursed RNA-Seq to comprehensively and quantitatively examine the dynamic transcriptional landscape of P. gingivalis in a model consortium with S. gordonii. Genes encoding a number of potential virulence determinants had higher relative mRNA levels in the context of dual species model communities than P. gingivalis alone, including adhesins, the Type IX secretion apparatus, and tetratricopeptide repeat (TPR) motif proteins. In contrast, genes encoding conjugation systems and many of the stress responses showed lower levels of expression in P. gingivalis. A notable exception to reduced abundance of stress response transcripts was the genes encoding components of the oxidative stress-related OxyR regulon, indicating an adaptation of P. gingivalis to detoxify peroxide produced by the streptococcus. Collectively, the results are consistent with evolutionary adaptation of P. gingivalis to a polymicrobial oral environment, one outcome of which is increased pathogenic potential.