Predicting the functional effects of missense variants is limited both by predictor performance and by the lack of principled strategies to integrate heterogeneous multiplexed assays of variant effect (MAVEs), including deep mutational scanning, into transferable sequence-function models. Here, we show that redundancy among partially overlapping deep mutational scanning experiments, often treated as noise, yields a consensus landscape of mutational constraint. We introduce “variant soundness”, an overlap-aware rank-fusion score that uses cross-experiment consistency to derive an assay-agnostic ordering of mutational tolerance within proteins. Applied to ~1100 MAVEs spanning over two million missense variants, this approach reveals a mutational landscape enriched for structural stability constraints. Guided by this integrated signal, we develop ESMRank, a sequence-based learning-to-rank predictor combining ESM2 representations with physicochemical descriptors. Under strict protein-level partitioning, ESMRank outperforms widely used predictors across abundance-, folding-, and stability-related assays in ProteinGym and Domainome. Predicted deleterious variants are enriched for ClinVar pathogenicity and, in CFTR, correlate with folding efficiency, channel activity, and pharmacological responsiveness. Together, these results turn assay redundancy into a scalable resource for interpretable missense variant prediction.

Point-of-care sensors have the potential to improve health equity in rural areas, where access to healthcare infrastructure and services is often limited. However, their readiness for scale-up and real-world deployment is judged mainly by device performance under controlled conditions. We argue that translational readiness should instead reflect how well a sensor fits the people, workflows, infrastructure and geography of its intended settings. A device that performs well in the laboratory may fail in rural settings when its design does not account for local operating conditions, workforce capacity and follow-up pathways. Here we introduce a geography-dependent readiness framework for translating point-of-care sensors into rural healthcare. The framework considers community priorities, usability, care-pathway integration and logistical constraints as upfront design criteria rather than downstream obstacles. Grounded in fieldwork across rural Australia, the framework could also apply to other regions where infrastructure, access and continuity of care are unevenly distributed. This Perspective presents a geography-dependent framework for assessing the readiness of point-of-care sensors for translation into rural healthcare settings, emphasizing the role of local context in guiding their design and deployment.

Methanogenic environments harbor complex prokaryotic consortia that drive the anaerobic conversion of organic matter into methane and carbon dioxide. Beyond being key to the global carbon cycle, such processes have been steered into powerful biotechnologies for energy production in bioreactors. Decoding the functional and taxonomic landscape of these communities requires comprehensive surveys, including bacteriophages and plasmids, which, however, remain vastly underrepresented in current databases. Here, we present the microbial Catalogue of Anaerobic Digestion (CAnDi), a database comprising 36 480 prokaryotic genomes spanning 11 149 species recovered from 1256 metagenomic samples collected across bioreactors worldwide. The reconstructed metagenomes also encode ∼1.9 million viral and over 130 000 plasmid sequences. Identity clustering revealed ∼536 000 candidate viral species, of which 85% are novel relative to existing databases, as well as over 15 000 unique plasmids. Clusters of viral and plasmid proteins were generated de novo, totalling 1 721 182 and 123 813 distinct protein families, respectively. Together, these resources substantially advance previous systematic approaches to characterize anaerobic digestion microbiomes, providing an expanded foundation for investigating the ecological and evolutionary roles of the virome and plasmidome in methane-producing ecosystems. CAnDi is accessible via https://www.candi-db.org/.

Homologous genes, especially those sharing close genomic locations and forming gene clusters, often have compensatory roles, therefore pose remarkable challenges for deciphering the functions and relationships of these genes. A comprehensive database of local homologous gene clusters with multimodal information would greatly facilitate such research. Here, we present LoGIC (https://omicslab.genetics.ac.cn/LoGIC/), a Local Gene-cluster Integrative Catalog that includes 20 306 local gene clusters identified from 30 representative eukaryotic species. In addition to the basic genomic features of each gene cluster, LoGIC also allows users to examine the sequence similarity of gene members within a cluster, the functional preference of clustered genes within a species, and the cross-species conservation of gene clusters. Information of tissue-wide expression profile and relationship with chromatin Hi-C structure is also available for human gene clusters. Together, LoGIC can facilitate studies of tandem gene duplication, paralog functional divergence, lineage-specific gene-family expansion, and evolutionary conservation or remodelling of homologous gene clusters across-species.

Antibiotic resistance is a global health threat requiring antimicrobials with alternative modes of action. Growing evidence indicates that bacteria organize key intracellular processes through biomolecular condensates and other forms of spatial organization, yet whether such spatial organization can be exploited as an antimicrobial target remains unclear. Here, we identify intrinsically disordered antimicrobial peptides, generated from the Drosophila sleep-inducing protein Nemuri, that inhibit bacterial growth by targeting ribosomal organization. Mechanistically, these peptides enter bacterial cells without overt membrane disruption and accumulate in ribosome-rich regions, where they bind broadly to RNA and ribosomal surfaces through multivalent interactions. This engagement remodels ribosome-rich regions into condensates with reduced ribosomal mobility, thereby suppressing global protein synthesis and distinguishing these peptides from canonical site-specific ribosome-targeting antibiotics. Furthermore, Nemuri-derived peptides exhibit broad-spectrum antimicrobial activity, including activity against drug-resistant pathogens, with low cytotoxicity toward Drosophila and mammalian cells. Together, our findings identify the organization and material dynamics of bacterial ribosomes as a biophysical vulnerability that can be exploited by antimicrobial peptides, revealing how a naturally evolved metazoan host-defense system turns intracellular organization itself into an antibacterial target. Nishinami et al. show that peptides from the sleep-inducing protein Nemuri, condense bacterial ribosomes and inhibit protein synthesis, revealing a biophysical antimicrobial strategy targeting intracellular organization.

The accelerating global energy demand, coupled with diminishing fossil fuel reserves and growing carbon emissions, underscores the urgent need for sustainable energy solutions. Autotrophic microalgae are promising cell factories for converting CO2 into renewable biofuels. However, conventional microalgal biofuels typically depend on energy-intensive or nutrient-deficient regimes that suppress microalgal proliferation, resulting in a critical trade-off between biomass growth and lipid biosynthesis. Here, we report a separable living interface with microalgae (SLIM) platform that employs extracellular-matrix-mimicking hydrogel scaffolds to spatially organize microalgae and dynamically reconfigure their microenvironment, enabling spatiotemporal switching between biomass accumulation or lipid induction. The SLIM platform supports high-density growth of microalgae while facilitating precision substrate-driven metabolic regulation, enabling low-carbon operation and improved lipid production (8.89 g L−1). This work establishes an engineered living material that overcomes essential bottlenecks in microalgal biofuel production and offers a generalizable blueprint for biohybrid systems in advanced biomanufacturing. Balancing biomass growth and lipid accumulation remains a challenge in microalgal biofuel production. Here the authors report extracellular matrix-mimicking hydrogel scaffolds that dynamically reconfigure the microenvironment of spatially organized microalgae, allowing controlled switching between the two states.

Safe and effective gene delivery remains a central challenge for therapeutic applications. While non-viral and viral vectors have enabled substantial progress, their reliance on non-human components often triggers immune responses, limiting their use in chronic treatments. Here, we develop DeepDelivery, an artificial intelligence-driven platform to identify human proteins with scaffold potential for mRNA delivery. Proteome-scale screening nominated 512 candidates, and experimental validation show that 12 of 15 selected human proteins functioned as scaffold components that facilitated particle formation and mRNA delivery within a modular engineering framework. Notably, several tripartite motif (TRIM) family proteins, typically linked to antiviral responses, exhibit strong assembly and delivery activity. Model interpretability and perturbation analyses identify TRIM domain features that influenced prediction scores and guide TRIM25-specific domain engineering to improve delivery. Together, these results establish a framework for mining the human proteome for scaffold components of programmable mRNA delivery systems and for further exploring the design space of human-derived delivery architectures. Protein-based carriers could provide biocompatible options for mRNA delivery. Here, authors develop an AI-guided platform to identify and engineer human-derived protein scaffolds that assemble into particles and support mRNA delivery in cultured cells.

Filamentous fungi of the genus Aspergillus, because of their exceptional secretory capacity and metabolic versatility have been extensively applied in traditional food fermentation and industrial biotechnology for production of diverse bioproducts. Their history of domestication and industrial application has enabled major progress in enzyme production; however, challenges associated with the filamentous growth and strain engineering for future biomanufacturing development still remain. This review summarizes the advances across the spectrum of strain-improvement methodologies, from classical mutagenesis, protoplast fusion, genome shuffling, and adaptive laboratory evolution to the modern era of CRISPR enabled genome editing, synthetic biology, and systems-level metabolic engineering. Previous reviews have primarily focused on individual Aspergillus species, specific industrial products or particular strain improvement approaches. In this review, we provide a comprehensive overview of classical and modern strain engineering technologies and industrial scale up across a wide range of Aspergillus species for industrial enzyme production. In particular, emerging non-GMO evolution tools such as atmospheric and room temperature plasma (ARTP) mutagenesis, and modern molecular biotechnology coupled with microdroplet-based ultrahigh-throughput screening, now allow the researchers to generate and screen a large number of Aspergillus species mutants with single-cell precision. The discovery of high-performance strains with enhanced secretion, stress tolerance, substrate utilization, and favorable enzymatic profiles for food and bioprocessing applications has been accelerated by the strain innovations. Collectively, these advances establish A. oryzae and the related species as increasingly reliable, scalable, and programmable hosts that can meet the biomanufacturing requirements of future circular and sustainable bioeconomies.

Biomolecular condensates have so far been studied in terms of their structural, compositional, and functional properties. However, condensate enzymatic activity - a key aspect of cellular metabolism - remains unexplored due to the complexity of the system. In this study, using a combination of experimental, computational and theoretical techniques, we have discovered that the non-equilibrium activity which originates from catalytic reactions couples with the environment through various feedback mechanisms across five orders of magnitude of length scales. We observe that condensed enzymes catalyze more rapidly in the presence of crowding proteins and show that the increased enzymatic activity within these droplets stems from the emergence of lower-energy protein conformations induced by the highly crowded environment. Despite the crowding in the environment of the droplet, which might suggest an effective increase in its overall viscosity, we find that it becomes more agile, as evidenced by the observation of enhanced diffusion and macroscopic flow, due to the enzymatic activity. These findings shed light on the dynamic interplay between enzymatic activity, composition and crowding in condensates, and their roles on the mobility and accessibility of various functional units in these environments, expanding our understanding of liquid-liquid phase separation in metabolically active conditions. Enzymes operate in highly crowded cellular environments, yet the impact of crowding on their function remains poorly understood. Here, the authors demonstrate that enzyme activity increases molecular mobility in crowded protein droplets, suggesting a feedback mechanism between catalysis and the physical properties of the surrounding environment.

Although machine learning has transformed protein structure prediction of folded protein ground states with remarkable accuracy, intrinsically disordered proteins and regions (IDPs/IDRs) are defined by diverse and dynamical structural ensembles that are predicted with low confidence by algorithms such as AlphaFold and RoseTTAFold. We present a machine learning method, IDPForge (Intrinsically Disordered Protein, FOlded and disordered Region GEnerator), that exploits a transformer protein language diffusion model to create all-atom IDP ensembles and IDR ensembles that maintains the folded domains. IDPForge does not require sequence-specific training, back transformations from coarse-grained representations, nor ensemble reweighting. The created IDP/IDR conformational ensembles show good agreement with both NMR and SAXS solution experimental data, and options for biasing with experimental restraints are available if desired. IDPForge with these diverse capabilities facilitates integrative and structural studies for proteins that contain intrinsic disorder, and is available as an open source resource for general use. Intrinsically disordered proteins and regions (IDPs/IDRs) are defined by diverse structural ensembles. We present IDPForge that exploits a transformer protein language diffusion model to create all-atom IDP and IDR ensembles that maintain folded domains with options to conform to experimental data.