Anaerobic fermentation with microbial communities (microbiomes) is an emerging platform for conversion of lignocellulosic biomass to biofuels and bioproducts. The process relies on diverse anaerobic microbes that interact to deconstruct and convert lignocellulosic biomass into a range of products, such as volatile fatty acids (VFAs), which can be achieved by arresting methanogenesis during fermentation. However, defining the distinct functional roles played by various fungi and bacteria during anaerobic biodegradation remains poorly understood. Here, we performed parallel enrichment experiments from cow faeces, goat faeces, and anaerobic digester sludge, selecting for fungal or bacterial dominated communities that convert sorghum biomass into VFAs. Subsequently we reconstructed metabolic networks across these enrichments based on recovered bacterial metagenome-assembled genomes (MAGs) and fungal isolate genomes and profiled their metabolic activity using metatranscriptomics to identify potential functional niches. Our findings implicate diverse bacteria affiliated with the Bacteroidales and Lachnospiraceae in the direct conversion of lignocellulosic biomass to propionate and butyrate, respectively, whereas Neocallimastix -dominated fungal enrichments converted lignocellulose to lactate, acetate and formate. Analysis of carbohydrate-active enzymes (CAZymes) revealed fine-scale differences between microbes that expressed unique multi-functional enzymes linking two or more CAZymes together with distinct carbohydrate binding motifs, implicating lignocellulose structure as a key driver of selection and niche differentiation. Most of these multi-functional enzymes localized complementary degradation functions together, likely conferring synergistic degradation effects within and between microbiome members. We anticipate that these findings will help inform efforts to develop synthetic microbiomes with tailored functionality for low-cost conversion of lignocellulosic biomass to fuels and bio-based chemicals. ### Competing Interest Statement The authors have declared no competing interest. Office of Biological and Environmental Research, https://ror.org/0114b2m14, DE-AC02-05CH11231 Office of Science, DE-SC0022142

The design of pathways to synthesize valuable molecules remains a central challenge in chemistry and biotechnology. Several computational retrosynthesis tools have been developed to address this problem, but their scope is often confined only to reactions in either synthetic organic chemistry or monofunctional enzymatic chemistry. We present TridentSynth, a web-based retrosynthesis tool (https://tridentsynth.lbl.gov) to scale synthesis planning up to three different routes by also incorporating multifunctional Type I polyketide synthase (PKS) enzymes into our reaction toolkit along with organic chemistry and monofunctional enzymes. Unlike monofunctional enzymes that catalyze single transformations, PKSs function as molecular assembly lines that catalyze multiple carbon-carbon bond formation reactions between acyl-coenzyme A substrates to construct elongated carbon scaffolds. PKSs follow a modular, programmable logic that allows them to be reconfigured to make new molecules in a predictable way. These scaffolds can then be chemoenzymatically modified to eventually access a wider array of molecular targets than would be possible with just synthetic chemistry or monofunctional enzymes alone, in a manner that mimics the evolved biosynthesis routes of many useful natural products. TridentSynth assists synthetic biologists by suggesting routes to synthesize a desired molecule through an intuitive web interface that requires no local installation or programming expertise.

CRISPR interference (CRISPRi) has emerged as a valuable tool for redirecting metabolic flux to enhance bioproduction. However, its application is often constrained by two challenges: (i) rationally identifying effective gene targets for downregulation and (ii) efficiently constructing multiplexed CRISPRi systems. In this study, we address both challenges by integrating a computational prioritization tool with a versatile assembly method for building multiplexed CRISPRi systems. FluxRETAP (Flux-Reaction Target Prioritization) accurately identified gene targets whose knockdown led to substantial increase of isoprenol titers in Pseudomonas putida KT2440, outperforming a conventional non-computational, pathway-guided target selection. The highest isoprenol titer of nearly 1.5 g/L was achieved by knocking down PP_4118 (a gene encoding α-ketoglutarate dehydrogenase). The use of VAMMPIRE (Versatile Assembly Method for MultiPlexing CRISPRi-mediated downREgulation) enabled accurate assembly of CRISPRi constructs containing up to five sgRNA arrays, reducing context dependency and achieving uniform, position-independent gene downregulation. The integration of FluxRETAP and VAMMPIRE has the potential to advance metabolic engineering by rapidly identifying CRISPRi-mediated knockdowns and knockdown combinations that enhance bioproduction titers, with potential applicability to other microbial systems.

Recent advances in artificial intelligence (AI) have rapidly changed the lab automation landscape, promoting self-driving laboratories (SDLs) that enable autonomous scientific discovery. These trends are increasingly applied in bioprocess development, yet bioprocessing faces unique challenges - biological complexity, regulatory and safety requirements, and multiscale experimentation - that distinguish it from other automation domains. Rather than pursuing full autonomy, we foresee that hybrid SDLs, combining AI-driven decision-making with sustained human oversight, represent the most practical near-term trajectory. This review examines three interconnected perspectives: (i) hybrid human-machine decision-making for bioprocessing; (ii) laboratory design considerations in the era of AI; and (iii) scale-up challenges when transitioning from screening to manufacturing. We highlight critical gaps in data standardization and the required community efforts necessary to realize autonomous bioprocess innovation.

Lipomyces starkeyi is an oleaginous yeast with a native metabolism well-suited for production of lipids and biofuels from complex lignocellulosic and waste feedstocks. Recent advances in genetic engineering tools have facilitated the development of L. starkeyi into a microbial chassis for biofuel and chemical production. However, the feasibility of redirecting L. starkeyi lipid flux away from lipids and towards other products remains relatively unexplored. Here, we engineer the native metabolism to produce malic acid by introducing the reductive TCA pathway and a C4-dicarboxylic acid transporter to the yeast. Heterogeneous expression of two genes, the Aspergillus oryzae malate transporter and malate dehydrogenase, enabled L. starkeyi malic acid production. Overexpression of a third gene, the native pyruvate carboxylase, allowed titers to reach approximately 10 g/L during shaking flasks cultivations, with production of malic acid inhibited at pH values less than 4. Corn-stover hydrolysates were found to be well-tolerated, and controlled bioreactor fermentations on the real hydrolysate produced 26.5 g/L of malic acid. Proteomic, transcriptomic and metabolomic data from real and mock hydrolysate fermentations indicated increased levels of a S. cerevisiae hsp9/hsp12 homolog (proteinID: 101453), glutathione dependent formaldehyde dehydrogenases (proteinIDs: 2047, 278215), oxidoreductases, and expression of efflux pumps and permeases during growth on the real hydrolysate. Simultaneously, machine learning based medium optimization improved production dynamics by 18

Synthetic biology’s remarkable potential to tackle important societal problems is held back by our inability to predictably engineer biological systems. Here, we collected one of the largest public multiomics synthetic biology datasets generated to date, and used it to train a novel deep learning algorithm able to predict product and metabolic dynamics with great accuracy, starting to approach the predictive capabilities found in physics and chemistry. We were able to predict production time series with 90-99% accuracy, and final production with 96% accuracy. Further, we were able to produce good predictions for a majority of extracellular metabolites, and twenty different intracellular metabolites. These predictions were provided for a target of industrial relevance: a non-model yeast ( Pichia kudriavzevii ) engineered to produce large amounts of malonic acid, a desirable biomanufacturing target. This approach is generally applicable to any host, pathway, and product because all required knowledge is inferred from experimental data. ### Competing Interest Statement H.G.M. declares financial interests in XLSI bio. United States Department of Energy, DE-EE0008489, DE-AC02-05CH11231, DE-NA0003525 Basque Government, BERC 2022–2025 Spanish Ministry of Science and Innovation, CEX2021-001142-S

Electroporation-mediated gene delivery is a cornerstone of synthetic biology, offering several advantages over other methods: higher efficiencies, broader applicability, and simpler sample preparation. Yet, electroporation protocols are often challenging to integrate into highly multiplexed workflows, owing to limitations in their scalability and tunability. These challenges ultimately increase the time and cost per transformation. As a result, rapidly screening genetic libraries, exploring combinatorial designs, or optimizing electroporation parameters requires extensive iterations, consuming large quantities of expensive custom-made DNA and cell lines or primary cells. To address these limitations, we have developed a High-Throughput Microfluidic Electroporation (HTME) platform that includes a 384-well electroporation plate (E-Plate) and control electronics capable of rapidly electroporating all wells in under a minute with individual control of each well. Fabricated using scalable and cost-effective printed-circuit-board (PCB) technology, the E-Plate significantly reduces consumable costs and reagent consumption by operating on nano to microliter volumes. Furthermore, individually addressable wells facilitate rapid exploration of large sets of experimental conditions to optimize electroporation for different cell types and plasmid concentrations/types. Use of the standard 384-well footprint makes the platform easily integrable into automated workflows, thereby enabling end-to-end automation. We demonstrate transformation of E. coli with pUC19 to validate the HTME’s core functionality, achieving at least a single colony forming unit in more than 99% of wells and confirming the platform’s ability to rapidly perform hundreds of electroporations with customizable conditions. This work highlights the HTME’s potential to significantly accelerate synthetic biology Design-Build-Test-Learn (DBTL) cycles by mitigating the transformation/transfection bottleneck.

Engineered polyketide synthases (PKSs) have great potential as biocatalysts. These unnatural enzymes are capable of synthesizing molecules that are either not amenable to biosynthesis or are extremely challenging to access chemically. PKSs can thus be a powerful platform to expand the chemical landscape beyond the limits of conventional metabolic engineering. Here we employ a retrobiosynthesis approach to design and construct PKSs to produce δ-valerolactam (VL) and three enantiopure α-substituted VL analogues that have no known biosynthetic route. We introduce the engineered PKSs and pathways for various malonyl-CoA derivatives into Pseudomonas putida and use proteomics, metabolomics and culture condition optimization to improve the production of our target compounds. These α-substituted VLs are polymerized into polyamides (nylon-5) or converted into their N-acryloyl derivatives. RAFT polymerization produces bio-derived polymers with potential biomedical applications. Overall, this interdisciplinary effort highlights the versatility and effectiveness of a PKS-based retrobiosynthesis approach in exploring and developing innovative biomaterials. Engineered polyketide synthases (PKSs) have great potential as biocatalysts for the synthesis of chemically challenging molecules. Here the authors show a retrobiosynthesis approach to design and construct PKSs to produce a series of valerolactams for biopolymer production.

MOTIVATION:Metabolic engineering is rapidly evolving as a result of new advances in synthetic biology tools and automation platforms that enable high throughput strain construction, as well as the development of machine learning tools (ML) for biology. However, selecting genetic engineering targets that effectively guide the metabolic engineering process is still challenging. ML can provide predictive power for synthetic biology, but current technical limitations prevent the independent use of ML approaches without previous biological knowledge. RESULTS:Here, we present FluxRETAP, a simple and computationally inexpensive method that leverages the prior mechanistic knowledge embedded in genome-scale models for suggesting targets for genetic overexpression, downregulation or deletion, with the final goal of increasing the production of a desired metabolite. This method can provide a list of desirable engineering targets that can be combined with current ML pipelines. FluxRETAP captured 100% of reaction targets experimentally verified to improve Escherichia coli isoprenol production, 50% of targets that experimentally improved taxadiene production in E. coli and ∼60% of genetic targets from a verified minimal constrained cut-set in Pseudomonas putida, while providing additional high priority targets that could be tested. Overall, FluxRETAP is an efficient algorithm for identifying a prioritized list of testable genetic and reaction targets. AVAILABILITY AND IMPLEMENTATION:FluxRETAP is implemented in python and released under the creative commons license. The implementation and code are freely available at: https://github.com/JBEI/FluxRETAP.

Although synthetic biology can produce valuable chemicals in a renewable manner, its progress is still hindered by a lack of predictive capabilities. Media optimization is a critical, and often overlooked, process which is essential to obtain the titers, rates and yields needed for commercial viability. Here, we present a molecule- and host-agnostic active learning process for media optimization that is enabled by a fast and highly repeatable semi-automated pipeline. Its application yielded 60% and 70% increases in titer, and 350% increase in process yield in three different campaigns for flaviolin production in Pseudomonas putida KT2440. Explainable Artificial Intelligence techniques pinpointed that, surprisingly, common salt (NaCl) is the most important component influencing production. The optimal salt concentration is very high, comparable to seawater and close to the limits that P. putida can tolerate. The availability of fast Design-Build-Test-Learn (DBTL) cycles allowed us to show that performance improvements for active learning are rarely monotonous. This work illustrates how machine learning and automation can change the paradigm of current synthetic biology research to make it more effective and informative, and suggests a cost-effective and underexploited strategy to facilitate the high titers, rates and yields essential for commercial viability.

Synthetic biology offers the promise of manufacturing chemicals more sustainably than petrochemistry. Yet, both the rate at which biomanufacturing can synthesize these molecules and the net chemical accessible space are limited by existing pathway discovery methods, which can often rely on arduous literature searches. Here, we introduce BioPKS pipeline, an automated retrobiosynthesis tool combining multifunctional type I polyketide synthases (PKSs) and monofunctional enzymes via two complementary tools: RetroTide and DORAnet. Monofunctional enzymes are valuable for carefully decorating a substrate's carbon backbone while PKSs are unique in their ability to iteratively catalyze carbon-carbon bond formation reactions, thereby expanding carbon backbones in a predictable fashion. We evaluate the performance of BioPKS pipeline using a previously reported set of 155 biomanufacturing candidates, achieving exact synthetic designs for 93 compounds and generating chemically similar pathways for most remaining targets. Furthermore, BioPKS pipeline can propose pathways for the complex therapeutic natural products cryptofolione and basidalin.

CRISPR interference (CRISPRi) has emerged as a valuable tool for redirecting metabolic flux to enhance bioproduction. However, conventional approaches for identifying target genes for CRISPRi-mediated downregulation have largely relied on heuristic methods and trial and error, which are labor-intensive and time-consuming. Additionally, the ability to achieve multigene knockdowns is limited by the constraints of molecular cloning techniques required for building multiplexed CRISPRi systems. In this study, we describe two novel methodologies FluxRETAP (Flux-REaction TArget Prioritization, a Genome-Scale Modeling Technique) and VAMMPIRE (a Versatile Assembly Method for MultiPlexing CRISPRi-mediated downREgulation). FluxRETAP accurately identified gene targets whose knockdown led to substantial increase of isoprenol titers, outperforming traditional heuristic selection. The use of VAMMPIRE enabled accurate and position-independent assembly of CRISPRi constructs containing up to five sgRNA arrays. The integration of FluxRETAP and VAMMPIRE has the potential to advance metabolic engineering by rapidly identifying CRISPRi-mediated knockdowns and knockdown combinations that enhance bioproduction titers, with potential applicability to other microbial systems. ### Competing Interest Statement The authors have declared no competing interest. United States Department of Energy, https://ror.org/01bj3aw27, DE-AC0205CH11231

Advances in genome engineering have improved our ability to perturb microbial metabolic networks, yet bioproduction campaigns often struggle with parsing complex metabolic datasets to efficiently enhance product titers. We address this challenge by coupling laboratory automation with machine learning to systematically optimize the production of isoprenol, a sustainable aviation fuel precursor, in Pseudomonas putida. The simultaneous downregulation through CRISPR interference of combinations of up to four gene targets, guided by machine learning, permitted us to increase isoprenol titer 5-fold in six consecutive design-build-test-learn cycles. Moreover, machine learning enabled us to swiftly explore a vast experimental design space of 800,000 possible combinations by strategically recommending approximately 400 priority constructs. High-throughput proteomics allowed us to validate CRISPRi downregulation and identify biological mechanisms driving production increases. Our work demonstrates that ML-driven automated design-build-test-learn cycles, when combined with rigorous data validation, can rapidly enhance titers without specific biological knowledge, suggesting that it can be applied to any host, product, or pathway. Laboratory automation, machine learning, and metabolic engineering may be combined to quickly and efficiently build productive microbial strains. Here the authors used these techniques in P. putida to boost isoprenol titers 5-fold over six DBTL cycles while sampling a reduced design space.

Sustainable aviation fuel (SAF) will significantly impact global warming in the aviation sector, and important SAF targets are emerging. Isoprenol is a precursor for a promising SAF compound DMCO (1,4-dimethylcyclooctane), and has been produced in several engineered microorganisms. Recently, Pseudomonas putida has gained interest as a future host for isoprenol bioproduction as it can utilize carbon sources from inexpensive plant biomass. Here, we engineer metabolically versatile host P. putida for isoprenol production. We employ two computational modeling approaches (Bilevel optimization and Constrained Minimal Cut Sets) to predict gene knockout targets and optimize the “IPP-bypass” pathway in P. putida to maximize isoprenol production. Altogether, the highest isoprenol production titer from P. putida was achieved at 3.5 g/L under fed-batch conditions. This combination of computational modeling and strain engineering on P. putida for an advanced biofuels production has vital significance in enabling a bioproduction process that can use renewable carbon streams.

Synthetic biology offers the promise of manufacturing chemicals more sustainably than petrochemistry. Yet, both the rate at which biomanufacturing can synthesize these molecules and the net chemical accessible space are limited by existing pathway discovery methods which rely on arduous literature searches. Here, we present an automated retrobiosynthesis tool, Biosynth Pipeline, that simultaneously tackles both problems by integrating multifunctional type I polyketide synthases (PKSs) with monofunctional enzymes to propose the synthesis of desired target chemicals via two new tools: DORAnet and RetroTide. While monofunctional enzymes are valuable for carefully decorating a substrate's carbon backbone, they typically cannot expand the backbone itself. PKSs can, instead, predictably do this through their unique ability to catalyze carbon-carbon bond formation reactions iteratively. We evaluated the performance of Biosynth Pipeline against a previously published set of 155 molecules of interest for biomanufacturing, and report that Biosynth Pipeline could produce exact designs for 93 of them, as well as pipelines to a chemically similar product for most of the remaining molecules. Furthermore, Biosynth Pipeline successfully proposes biosynthetic routes for complex therapeutic natural products (cryptofolione and basidalin) for which no known biosynthetic pathway currently exists. ### Competing Interest Statement J.D.K. has financial interests in Amyris, Ansa Biotechnologies, Apertor Pharma, Berkeley Yeast, Cyklos Materials, Demetrix, Lygos, Napigen, ResVita Bio and Zero Acre Farms. The other authors declare no competing interests.

Engineering metabolism to efficiently produce chemicals from multi-step pathways requires optimizing multi-gene expression programs to achieve enzyme balance. CRISPR-Cas transcriptional control systems are emerging as important tools for programming multi-gene expression, but poor predictability of guide RNA folding can disrupt expression control. Here, we correlate efficacy of modified guide RNAs (scRNAs) for CRISPR activation (CRISPRa) in E. coli with a computational kinetic parameter describing scRNA folding rate into the active structure (rS = 0.8). This parameter also enables forward design of scRNAs, allowing us to design a system of three synthetic CRISPRa promoters that can orthogonally activate (>35-fold) expression of chosen outputs. Through combinatorial activation tuning, we profile a three-dimensional design space expressing two different biosynthetic pathways, demonstrating variable production of pteridine and human milk oligosaccharide products. This RNA design approach aids combinatorial optimization of metabolic pathways and may accelerate routine design of effective multi-gene regulation programs in bacterial hosts. Guide RNA folding affects functionality of CRISPR-Cas transcriptional control systems. Here, the authors report computational gRNA design together with creation of synthetic CRISPRa promoters for orthogonal expression control and demonstrate the application in pteridine and human milk oligosaccharide production in E. coli.

ADVERTISEMENT RETURN TO ISSUEEditorialNEXTSpecial Issue on Artificial Intelligence for Synthetic BiologyHector García Martín*Hector García MartínBiological Systems and Engineering Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United StatesBCAM − Basque Center for Applied Mathematics, Bilbao 48009, SpainDOE Agile BioFoundry, Emeryville, California 94608, United StatesDOE Joint BioEnergy Institute, Emeryville, California 94608, United States*Email: [email protected]More by Hector García Martínhttps://orcid.org/0000-0002-4556-9685, Stanislav Mazurenko*Stanislav MazurenkoLoschmidt Laboratories, Department of Experimental Biology and RECETOX, Faculty of Science, Masaryk University, Kamenice 5, 625 00 Brno, Czech RepublicInternational Clinical Research Center, St. Anne's University Hospital Brno, Pekarska 53, 656 91 Brno, Czech Republic*Email: [email protected]More by Stanislav Mazurenko, and Huimin Zhao*Huimin ZhaoDepartment of Chemical and Biomolecular Engineering, University of Illinois at Urbana−Champaign, Urbana, Illinois 61801, United States*Email: [email protected]More by Huimin Zhaohttps://orcid.org/0000-0002-9069-6739Cite this: ACS Synth. Biol. 2024, 13, 2, 408–410Publication Date (Web):January 12, 2024Publication History Published online12 January 2024Published inissue 16 February 2024https://pubs.acs.org/doi/10.1021/acssynbio.3c00760https://doi.org/10.1021/acssynbio.3c00760editorialACS PublicationsCopyright © Published 2024 by American Chemical Society. This publication is available under these Terms of Use. Request reuse permissions This publication is free to access through this site. Learn MoreArticle Views5097Altmetric-Citations-LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InRedditEmail PDF (1 MB) Get e-AlertscloseSUBJECTS:Genetics,Kinetic modeling,Optimization,Peptides and proteins,Synthetic biology Get e-Alerts

Sustainably grown biomass is a promising alternative to produce fuels and chemicals and reduce the dependency on fossil energy sources. However, the efficient conversion of lignocellulosic biomass into biofuels and bioproducts often requires extensive testing of components and reaction conditions used in the pretreatment, saccharification, and bioconversion steps. This restriction can result in a significant and unwieldy number of combinations of biomass types, solvents, microbial strains, and operational parameters that need to be characterized, turning these efforts into a daunting and time-consuming task. Here we developed a high-throughput feedstocks-to-fuels screening platform to address these challenges. The result is a miniaturized semi-automated platform that leverages the capabilities of a solid handling robot, a liquid handling robot, analytical instruments, and a centralized data repository, adapted to operate as an ionic-liquid-based biomass conversion pipeline. The pipeline was tested by using sorghum as feedstock, the biocompatible ionic liquid cholinium phosphate as pretreatment solvent, a “one-pot” process configuration that does not require ionic liquid removal after pretreatment, and an engineered strain of the yeast Rhodosporidium toruloides that produces the jet-fuel precursor bisabolene as a conversion microbe. By the simultaneous processing of 48 samples, we show that this configuration and reaction conditions result in sugar yields (~70%) and bisabolene titers (~1500 mg/L) that are comparable to the efficiencies observed at larger scales but require only a fraction of the time. We expect that this Feedstocks-to-Fuels pipeline will become an effective tool to screen thousands of bioenergy crop and feedstock samples and assist process optimization efforts and the development of predictive deconstruction approaches.

AbstractMetabolic fluxes, the number of metabolites traversing each biochemical reaction in a cell per unit time, are crucial for assessing and understanding cell function.13C Metabolic Flux Analysis (13C MFA) is considered to be the gold standard for measuring metabolic fluxes.13C MFA typically works by leveraging extracellular exchange fluxes as well as data from13C labeling experiments to calculate the flux profile which best fit the data for a small, central carbon, metabolic model. However, the nonlinear nature of the13C MFA fitting procedure means that several flux profiles fit the experimental data within the experimental error, and traditional optimization methods offer only a partial or skewed picture, especially in “non-gaussian” situations where multiple very distinct flux regions fit the data equally well. Here, we present a method for flux space sampling through Bayesian inference (BayFlux), that identifies the full distribution of fluxes compatible with experimental data for a comprehensive genome-scale model. This Bayesian approach allows us to accurately quantify uncertainty in calculated fluxes. We also find that, surprisingly, the genome-scale model of metabolism produces narrower flux distributions (reduced uncertainty) than the small core metabolic models traditionally used in13C MFA. The different results for some reactions when using genome-scale models vs core metabolic models advise caution in assuming strong inferences from13C MFA since the results may depend significantly on the completeness of the model used. Based on BayFlux, we developed and evaluated novel methods (P-13C MOMA and ROOM) to predict the biological results of a gene knockout, that improve on the traditional MOMA and ROOM methods. We provide an open source Python implementation of BayFlux athttps://github.com/JBEI/bayflux.Author summary13C MFA practitioners know that modeling results can be sensitive to minor modifications of the metabolic model. Certain parts of the metabolic model that are not well mapped to a molecular mechanism (e.g.drains to biomass or ATP maintenance) can have an inordinate impact on the final fluxes. The only way to ascertain the validity of the model is by checking that the result does not significantly differ from previously observed flux profiles. However, that approach diminishes the possibility of discovering truly novel flux profiles. Because of this strong dependence on metabolic model details, it would be very useful to have a systematic and repeatable way to produce these metabolic models. And indeed there is one: genome-scale metabolic models can be systematically obtained from genomic sequences, and represent all the known genomically encoded metabolic information. However, these models are much larger than the traditionally used central carbon metabolism models. Hence, the number of degrees of freedom of the model (fluxes) significantly exceeds the number of measurements (metabolite labeling profiles and exchange fluxes). As a result, one expects many flux profiles compatible with the experimental data. The best way to represent these is by identifying all fluxes compatible with the experimental data. Our novel method BayFlux, based on Bayesian inference and Markov Chain Monte Carlo sampling, provides this capability. Interestingly, this approach leads to the observation that traditional optimization approaches can significantly overestimate flux uncertainty, and that genome-scale models of metabolism produce narrower flux distributions than the small core metabolic models that are traditionally used in13C MFA. Furthermore, we show that the extra information provided by this approach allows us to improve knockout predictions, compared to traditional methods. Although the method scales well with more reactions, improvements will be needed to tackle the large metabolic models found in microbiomes and human metabolism.

Human exploration of deep space will involve missions of substantial distance and duration. To effectively mitigate health hazards, paradigm shifts in astronaut health systems are necessary to enable Earth-independent healthcare, rather than Earth-reliant. Here we present a summary of decadal recommendations from a workshop organized by NASA on artificial intelligence, machine learning and modelling applications that offer key solutions toward these space health challenges. The workshop recommended various biomonitoring approaches, biomarker science, spacecraft/habitat hardware, intelligent software and streamlined data management tools in need of development and integration to enable humanity to thrive in deep space. Participants recommended that these components culminate in a maximally automated, autonomous and intelligent Precision Space Health system, to monitor, aggregate and assess biomedical statuses. Deep-space exploration missions require new technologies that can support astronaut health systems as well as biological monitoring and research systems that can function independently from Earth-based mission control centres. A NASA workshop explored how artificial intelligence advances could help address these challenges and, in this first of two Review articles based on the findings from the workshop, a vision for autonomous biomonitoring and precision space health is discussed.