AI co-scientists can act as virtual research collaborators in statistical genetics, accelerating genetic discovery and translation. Realizing this potential depends on robust domain-specific data and infrastructures, together with interdisciplinary teamwork to build community standards that ensure rigor and responsible deployment.

Bacteriophages have evolved diverse strategies to manipulate host processes, yet the molecular mechanisms employed by phage-encoded effector proteins remain poorly understood. Here, we identify Gp16, an early-expressed protein from Staphylococcus aureus phage ΦNM1, as a RecA-dependent nuclease that plays a dual role in host inhibition and phage propagation. Gp16 is rapidly expressed upon infection, and its overexpression alone is sufficient to inhibit bacterial growth where deletion of gp16 severely impairs phage DNA replication, progeny production, and host cell lysis, underscoring its essential role in the phage life cycle. Structural modeling predicts Gp16 is a nuclease, and its overexpression induces DNA condensation in vivo. Biochemical and cellular analyses show that Gp16 interacts with the host RecA protein to inhibit growth, and functions as a nickase in vitro, requiring the catalytic cysteine C181 for DNA cleavage. RecA further enhances its cleavage activity. During phage infection, RecA activation is required for efficient phage propagation, while Gp16 concurrently suppresses host DNA replication and promotes DNA condensation, thereby facilitating phage replication. Together, these findings reveal a previously unrecognized strategy in which a phage-encoded nuclease exploits the host RecA machinery to couple host suppression with productive phage propagation.

Eukaryotic Fanzor proteins are compact, programmable RNA-guided nucleases with substantial potential for genome editing, although their efficiency in mammalian cells remains suboptimal. Here, we present a combinatorial engineering strategy to optimize a representative Fanzor system, MmeFz2–ωRNA. AlphaFold3-powered rational redesign produced a minimized ωRNA scaffold that is 30% smaller while maintaining up to 82.2% efficiency. Synergistic structure-guided and AI-augmented protein engineering generated two variants, enMmeFz2 and evoMmeFz2, which exhibited an average ~32-fold increase in activity across 38 genomic loci. Moreover, fusion of the non-specific DNA-binding domain HMG-D further enhanced editing performance (enMmeFz2-HMG-D and evoMmeFz2-HMG-D). Notably, evoMmeFz2-HMG-D demonstrated robust in vivo genome editing activity, enabling dystrophin restoration in humanized male Duchenne muscular dystrophy mouse models via single adeno-associated virus (AAV) delivery. This study establishes Fanzor2 as a gene editing platform for genome engineering and therapeutic applications, and underscores the power of AI-guided engineering to accelerate genome editor development while reducing experimental burden. Eukaryotic Fanzor proteins are compact and advantageous for delivery, but their activity remains limited. Here, the authors engineer an improved Fanzor2 system (evoMmeFz2) using structure-guided and AI-assisted strategies to enable efficient exon skipping in a Duchenne muscular dystrophy model.

Continuous cropping often exacerbates soil-borne diseases, particularly Fusarium wilt, yet the intricate rhizosphere relationships among phyto-derived metabolites, pathogens, and particular microbial functions remain poorly understood. Here, we observe that citrulline accumulation during continuous cropping is positively correlated with Fusarium wilt severity by enhancing fusaric acid production in Fusarium oxysporum. Metagenomic analyses reveal that citrulline turnover-related functions, represented by functional modules including M00978, are significantly enriched in healthy rhizosphere soils but are notably reduced in Fusarium-conducive soils. The functional genes, arcB and argH, are identified in Pseudomonas putida YDTA3, with arcB being essential for citrulline-degradation via knockout experiments. The inoculation of an arcB-expressing indigenous Escherichia consortium (EO-arcB) in three independent continuous cropping systems of cucurbit crops demonstrates that enhancing and maintaining the soil citrulline-degrading function mitigates soil-borne Fusarium wilt. In summary, this study advances our understanding of rhizosphere interactions underlying Fusarium wilt disease occurrence and offers a promising biocontrol strategy. In this study, the authors show that citrulline accumulation exacerbates Fusarium wilt and that enhancing the soil’s citrulline-degrading function, particularly via the arcB gene, significantly reduces disease in continuous cucurbitaceous cropping systems.

Bacteriophages are considered to have great potential as naturally occurring, antimicrobial agents for use in food production. Phages are ubiquitous in nature and can be isolated from almost all habitats. This review outlines the possibilities, as well as limitations of their use in food production. Application of phages in the food sector are described and the limitations of their use as well as potential risks are discussed. Approaches for a possible classification as either processing aid or food additive are considered, and the current status of their use in and outside the EU is presented. Finally, the need for research to close identified knowledge gaps is highlighted.

Fiber, the most abundant organic polymer in nature, is widely recognized as a foundational sustainable material with diverse applications across industrial, medical, and consumer domains. Owing to its renewability and widespread availability, it also serves as a critical alternative energy source in agriculture, enabling more sustainable livestock production through the efficient conversion of fibrous feedstuffs, thereby supporting the principles of a circular bioeconomy. Cellulose, which constitutes up to 80% of plant fiber, contains tightly packed crystalline regions that confer strong resistance to microbial degradation. Other key obstacles to efficient cellulose digestion in the gut include the absence of critical cellulolytic genes, low enzymatic activity, a lack of natural activators, and the presence of cellulase inhibitors. Synthetic biology provides innovative molecular-level strategies to overcome key technical barriers in cellulose degradation. These approaches employ targeted modifications at nucleic acid and protein levels, including the introduction of engineered genes, synthetic regulators, and optimized enzymes, to develop high-performance microbial systems with enhanced cellulose-degrading capabilities. Furthermore, genetic modifications like the knockout of inhibitory genes and knock-in of activator genes, combined with rational redesign of multi-enzyme complexes, can significantly improve the secretion and catalytic efficiency of cellulases. When integrated with artificial intelligence, synthetic biology enables predictive screening and precision engineering of microbial strains for highly efficient cellulose degradation. This review comprehensively summarizes recent advances in synthetic biology approaches for improving cellulose degradation and highlights how these tools can optimize fiber utilization in sustainable agricultural and industrial applications.

The nitrogen-fixing symbiosis between leguminous plants and soil bacteria, collectively termed rhizobia, is a major contributor of fixed nitrogen to the biosphere. The ability of legumes to secure nitrogen from the atmosphere underlies their ecological success and has made them important crops in both traditional and modern sustainable agriculture. Many genes directing the establishment and functioning of this beneficial interaction have been identified under laboratory conditions using a limited number of bacterial strains and plant species in pairwise combinations. Under natural and field conditions, however, plants encounter numerous potential partners, as soil microbiomes contain diverse bacteria equipped with the necessary toolkit for symbiosis. Consequently, legumes must possess mechanisms to select for or against specific partners. This review highlights how legumes employ elements of their immune system for the negative selection of rhizobia via processes resembling the gene-for-gene model of effector-triggered immunity in plant–pathogen interactions.

TenA (transcription enhancement) family proteins are widespread in bacteria, yet their functions in regulating natural product biosynthesis remain largely unexplored. Here, we report that LysR2, a TenA-family protein encoded by the lysolipin I biosynthetic gene cluster, acts as a transcriptional activator. Genetic and biochemical analyses reveal that LysR2 employs a noncanonical mechanism. Lacking a standard DNA-binding domain, it does not bind target promoters directly. Instead, LysR2 activates transcription by specifically binding to the pathway-specific repressor, LysR1. This protein–protein interaction displaces LysR1 from its cognate promoter, thereby derepressing the transcription of the biosynthetic gene cluster. Our work provides the first experimental evidence for a regulatory role of a TenA-family protein and elucidates a unique indirect activation mechanism that broadens the paradigm of transcriptional regulation in bacterial natural product biosynthesis.

Following their success in natural language processing and protein biology, pretrained large language models have started appearing in genomics in large numbers. These genomic language models (gLMs), trained on diverse DNA and RNA sequences, promise improved performance on a variety of downstream prediction and understanding tasks. In this review, we trace the rapid evolution of gLMs, analyze current trends, and offer an overview of their application in genomic research. We investigate each gLM component in detail, from training data curation to the architecture, and highlight the present trends of increasing model complexity. We review major benchmarking efforts, suggesting that no single model dominates, and that task-specific design and pretraining data often outweigh general model scale or architecture. In addition, we discuss requirements for making gLMs practically useful for genomic research. While several applications, ranging from genome annotation to DNA sequence generation, showcase the potential of gLMs, their use highlights gaps and pitfalls that remain unresolved. This guide aims to equip researchers with a grounded understanding of gLM capabilities, limitations, and best practices for their effective use in genomics.