CRISPR-Cas9 nucleases have transformed genome engineering, yet their application is often constrained by protospacer-adjacent motif (PAM) requirements. Staphylococcus aureus Cas9 (SaCas9) is particularly attractive for in vivo delivery due to its compact size, but its NNGRRT PAM limits targetable genomic sites. Here, we report KRH (E782K/N968R/R1015H), a SaCas9 variant designed entirely through an improved, fully computational protein point-mutation design workflow, UniDesign, without additional experimental optimization. KRH efficiently recognizes the expanded NNNRRT PAM, achieving genome- and base-editing efficiencies comparable to those of the evolution-derived KKH variant across multiple human cell types. Structural and energetic analyses reveal that KRH relaxes PAM specificity by fine-tuning the balance between sequence-specific interactions with PAM bases and nonspecific contacts with the DNA backbone. Beyond its practical utility, KRH demonstrates that computational design can identify a minimal set of mutations sufficient to remodel the PAM interface while preserving high nuclease activity. This approach not only recapitulates evolution-derived performance but also, in some cases, surpasses it, offering a scalable strategy for high-throughput Cas9 variant development. Overall, KRH establishes a blueprint for rationally engineered, PAM-relaxed nucleases and underscores the potential of computational protein design to accelerate next-generation genome editing, complementing traditional molecular evolution approaches.

Analyzing RNA structural ensembles, whether derived from molecular dynamics, enhanced sampling techniques, or experimental data, poses significant challenges due to the intrinsic flexibility and diverse conformational landscapes of RNA. We present ARNy Plotter, a web-based platform designed for comprehensive analysis and visualization of RNA structural ensembles. ARNy Plotter integrates multiple state-of-the-art methods into a unified, user-friendly environment, enabling intuitive examination of RMSD and eRMSD distributions, torsion-angle variability, dynamic secondary-structure patterns, contact-frequency networks, and multidimensional conformational landscapes. By leveraging MDAnalysis, Barnaba, and foRNA, the platform supports all major trajectory and ensemble formats while offering real-time 3D visualization via MolStar. Unlike existing tools that focus primarily on static structures or provide limited dynamic analysis, ARNy Plotter allows in-depth exploration of conformational transitions and population distributions within RNA ensembles, and facilitates result sharing without distributing full datasets. This platform addresses the growing need for accessible, comprehensive, and interactive tools for RNA structural biology. https://arny-plotter.rpbs.univ-paris-diderot.fr/ 

Accurate prediction of protein-ligand binding affinity is central to computational drug discovery. Recent foundation models, such as Boltz-2, have achieved remarkable accuracy, but their high computational cost poses a major barrier to large-scale virtual screening. We address this challenge by introducing a lightweight structure based virtual screening model, FlashAffinity, that achieves similar performance as Boltz-2 in affinity prediction and binder classification tasks, while achieving a 50x speedup at inference time. FlashAffinity replaces the expensive protein structure prediction models with a simple protein-ligand docking model and the PairFormer-based affinity scoring module with a cheap EGNN architecture. In summary, this work bridges the gap between accuracy and efficiency, enabling ultra-fast virtual screening of massive chemical libraries.

High-performance cell-free protein synthesis has transformative potential for synthetic biology, yet the prohibitive costs of PURE kits and the labor intensity of in-house preparation have restricted accessibility and scalability. We developed i-POPFLEX (Purified components OPtimized for FLEXible protein expression), a modular cell-free protein synthesis system in which 34 translation factors, individually synthesized in vitro, are assembled using automated liquid handling. This workflow minimizes manual input and supports parallelized production, generating complete, ready-to-use systems within two days. i-POPFLEX achieves up to 5-fold higher protein yields and a 95 % cost reduction (20-fold lower cost) compared with commercial kits. Its flexible architecture also enables selective component inclusion for genetic code reprogramming and site-specific incorporation of non-canonical amino acids. By coupling modular design with automation, i-POPFLEX provides an accessible, customizable, and economically viable platform for next-generation biomanufacturing workflows.

Two-stage bioprocesses which decouple cell growth from product synthesis are an attractive approach to biomanufacturing. However high levels of production in stationary phase cultures often suffer from a progressive decline in metabolism. We demonstrate that in E. coli pyruvate accumulation, an inevitable consequence of high-flux metabolism, acts as a major inhibitor of stationary-phase glucose uptake. We present a novel central metabolism to optimize stationary phase production, a gluconate-bypass, which circumvents this challenge by rerouting carbon flux around glycolysis. This redesign achieves two critical outcomes: first, it decouples glucose uptake from pyruvate inhibition; second, glucose oxidation intrinsically co-generates the NADPH cofactor. Validated using NADPH-dependent L-alanine as a representative model, the GBP creates a self-regulating host that achieved a record titer of 197 g/L with a 1.6-fold extension of production longevity. This work establishes the GBP as a generalizable platform for robust stationary phase biosynthesis.

Phage therapy, which uses viruses that infect bacteria to target and lyse specific bacterial pathogens, has re-emerged as a promising strategy to combat antibiotic-multiresistant bacteria. Advances in metagenomics and synthetic biology, together with systems biology approaches combining mathematical modeling with experimental data, provide excellent opportunities to understand phage-bacteria dynamics. Here we analyze a mathematical model successfully calibrated using in vitro data on the multidrug-resistant bacterium Klebsiella pneumoniae in the presence of the phage vB Kpn 2-P4. The model describes a system with a susceptible bacterial population that can generate phage-resistant mutants. By analyzing the equilibria and bifurcations of the model, we identify a coexistence scenario between phage-resistant bacteria and phages governed by a quasineutral line of equilibria with both stable and unstable segments. Biologically, this quasineutral structure implies that phage-resistant bacteria can persist across a wide range of phage densities without selective pressure favoring a unique outcome, making clearance highly sensitive to additional mortality mechanisms. The clearance of phage-resistant bacteria can be achieved by combining phage activity with an increased death rate of the resistant strains. This process is governed by a global transcritical bifurcation of the quasineutral line. Our model offers mechanistic insight into potential scenarios leading to the complete elimination of phage-resistant bacteria.

Bacillus subtilis is a critical host for protein production, with many industrial strains relying on strong constitutive promoters. However, this kind of promoter typically imposes a heavy burden on the host from the early stage of fermentation, leading to reduced growth rate and biomass. To overcome the drawbacks of these promoters, we developed a xylose-inducible CRISPRi module to dynamically control the activity of these promoters. The strength of this module was finely tuned via promoter engineering and the xylose concentration. The addition of xylose inhibited the target promoter and favored cell growth at an early stage, while the consumption of xylose recovered the strength of the promoter and facilitated protein expression, resulting in better balance between cell growth and protein production. The yield of a target protein was increased by 38% using this module. Our work provides a simple and effective method to upgrade industrial strains driven by strong constitutive promoters.

Isobutanol, a promising biofuel with higher energy content than ethanol, presents a sustainable alternative through biosynthesis. However, enhancing yield remains challenging due to the inefficiencies in microbial synthesis. This study introduces a transcription factor-based biosensor using the AlkS-PalkB system in Escherichia coli, which correlates green fluorescence with isobutanol concentration. Employing directed evolution, we modified AlkS to detect isobutanol, significantly improving biosensor specificity. Initial modifications increased the dynamic response from non-detectable to a 2.60-fold change. Subsequent optimizations through site-directed mutagenesis and promoter engineering further enhanced this response to a 5.56-fold change, equivalent to a 114% increase. Although engineered for isobutanol detection with high sensitivity, the engineered biosensor retains responsiveness to several short-chain alcohols. This biosensor provides a foundation for high-throughput screening of isobutanol and other short-chain alcohol-producing strains, though additional improvements in selectivity and operating range may be required for efficient implementation.

Protein-protein interactions (PPIs) are essential for the study of cellular function, yet computational prediction of bacterial PPIs remains limited. Most existing methods are trained on human data, reducing their applicability to bacterial systems. Here, we present B-PPI, a computational tool specifically designed for bacterial PPI prediction. B-PPI leverages embeddings from ProstT5, a structure-aware protein language model, and a cross-attention mechanism to capture residue-level inter-protein relationships. To facilitate training, we constructed B-PPI-DB, a large-scale bacterial PPI dataset derived from STRING, comprising 202,829 positive and negative interactions across 2,646 taxa with a 1:10 positive-to-negative ratio. We benchmarked B-PPI against TT3D, a state-of-the-art model trained on human PPI, which was previously evaluated on bacterial PPIs. B-PPI achieved substantially higher performance on bacterial data (AUPRC 0.926 vs. 0.230 and F1 0.866 vs. 0.299) with faster runtime. We further demonstrate that the model adapts to unseen bacterial interactions with minimal fine-tuning. Together, B-PPI and B-PPI-DB address a critical gap in computational microbiology, offering a framework for bacterial PPI prediction and a data resource for benchmarking and developing new tools in the field.

Over the last decade, the expansion in the number of available genomes has profoundly transformed the study of genetic diversity, evolution, and ecological adaptation in prokaryotes. However, traditional bioinformatic approaches based on the analysis of individual genomes are showing their limitations when faced with the sheer scale of the data. To overcome these constraints, the concept of pangenome has emerged, offering a comprehensive framework to capture the full genetic repertoire of a species. In this study, we present PANORAMA, an innovative pangenomic tool designed to exploit pangenome graphs and enable them to be annotated and compared in order to explore the genomic diversity of several species. Based on the PPanGGOLiN pangenome graphs, PANORAMA integrates advanced methods for rule-based prediction of macromolecular systems and comparative analysis of conserved features between different pangenomes, such as spots of insertion. We illustrate the use of PANORAMA on a dataset of 941 Pseudomonas aeruginosa genomes, evaluating its performance against reference defense system prediction tools such as PADLOC and DefenseFinder. The analysis was then extended to a larger set, including four species of Enterobacteriaceae (>6,000 genomes), demonstrating PANORAMA’s ability to annotate, compare, and explore the diversity and distribution of biological systems across multiple species. This work provides new methods for the large-scale comparative study of microbial genomes and underlines the relevance of pangenome approaches in deciphering their evolutionary dynamics. PANORAMA is freely available and accessible through: https://github.com/labgem/PANORAMA

Levan is a fructose polymer with applications in the creation of hydrogels for drug delivery and wound healing. In industrial biotechnology. Bacillus subtilis is the key organism for producing levan. However, the metabolic models of B. subtilis available do not include the biosynthesis of levan. To understand levan biosynthesis in B. subtilis, we employed structural systems biology integrating known structural details of proteins in the B. subtilis metabolic pathway to create a structure-annotated genome-scale metabolic model (GEM). To fill gaps in structural information about the enzymes, AlphaFold2 was used. Thus, this study enhances the metabolic model of B. subtilis by incorporating the synthesis of levan and including structural information about the proteins involved. The manually curated model links proteins and reactions to protein data bank (PDB) entries, providing structural perspectives previously overlooked in GEMs. We mapped 508 PDB structures to 168 UniProt IDs to unravel 331 out of 1250 reactions (26.5%) in B. subtilis with focused coverage of sacB, sacC, sacX/Y, levD/E/F/G, and sacP. The structural layer does not alter stoichiometry or constraints unless explicitly parameterized. This structure-annotated resource enables the systematic testing of phenotype predictions and design strategies. Our structure-based metabolic model advances the understanding of levan production and microbial metabolism, facilitating sustainable and efficient biotechnological processes for industrial applications.