RMH

Computational bacteriophage host prediction from genomic sequences remains challenging because host range depends on diverse, rapidly evolving genomic determinants—from receptor-binding proteins to anti-defense systems and downstream infection compatibility—and because the signals available to predictors, including sequence homology, CRISPR spacer matches, nucleotide composition, and mobile genetic elements, are sparse, unevenly distributed across taxa, and constrained by incomplete host annotations. Here, we frame host prediction as an unsupervised retrieval problem. We asked whether embeddings from the pretrained genome language model Evo2 captured a reliable host-range signal without training on phage–host labels. We generated whole-genome embeddings for phages and candidate bacterial hosts with the Evo2-7B model, applied normalization, and ranked hosts by cosine similarity. Using the Virus-Host Database, we selected embedding and fusion choices on a Gram-positive validation cohort and then evaluated the approach on a held-out Gram-negative test cohort to minimize data leakage. We found that Evo2 was strongest at retrieving multiple plausible hosts, with the recorded host in the top 10 for 55.4% of phages. However, it did not maximize species-level top-1 accuracy (19.4% vs. 23.2% for the best baseline). At higher taxonomic ranks, Evo2 captured a coarser host-range signal: top-1 accuracy reached 43.4% at the genus level and 51.6% at the family level. Reciprocal rank fusion of Evo2 with BLASTN, VirHostMatcher, and PHIST improved all retrieval metrics. Top-10 retrieval rose to 58.5% and top-1 accuracy to 26.9%. Stratified analyses by phage genome length, host clade, and host mobile genetic element coverage revealed scenario-dependent performance. Evo2 embeddings excelled for intermediate-length phages and when host mobile element content was low, whereas alignment and k-mer methods dominated when local homology was abundant. These results suggest that pretrained genome embeddings complement established alignment- and k-mer/composition-based methods and that context-aware hybrid pipelines may help improve phage host prediction.

Efficient production of human proteins for the development of tool compounds and biologics depends on a detailed understanding of the protein expression machinery in mammalian cells. Codon optimization is widely believed to enhance protein yield, yet its impact in homologous mammalian systems remains poorly defined. Here, we systematically compare five codon usage strategies reflecting common assumptions about rare codons, RNA stability, and synthesis efficiency. We developed pTipi, an efficient open source mammalian expression vector, and evaluated its performance in antibody production. We generated plasmids for common epitope tag antibodies such as V5, anti-biotin and anti-His for distribution by Addgene. To compare codon usage schemes, we performed a bake-off of 18 human and murine Wnt pathway glycoproteins in mammalian cells. Small-scale expression screens revealed that codon optimization does not provide a general advantage over native coding sequences, while strategies prioritizing RNA stability consistently reduced expression. Interestingly, a skewed codon scheme using the most abundant codons produced yields comparable to native sequences and occasionally enhanced protein output. To enable flexible evaluation of codon strategies, we implemented a Golden Gate compatible pTipi platform for efficient synthetic gene incorporation. We conclude that native codons are sufficient for robust homologous mammalian expression of glycoproteins, while selective codon skewing can be beneficial for some targets. 

In recent years, CRISPR/Cas gene editing technology has become a fundamental method in biological breeding. As a vital tool for overcoming technological obstacles, it is currently widely used in functional gene research and genetic enhancement across a variety of organisms. Currently, CRISPR activation (CRISPRa) technology based on dCas9 fusion transcription activation domains has emerged as a powerful tool for expanding the application of CRISPR/Cas systems in improving traits in plants, animals, and microorganisms. This overview starts by going over the underlying principles and components of gene activation editing technology, as well as the phases of development of its three generations. It summarises the present difficulties and potential directions in this field while concentrating on the use of gene activation editing in important crop traits including growth and development regulation, stress resistance, and quality regulation. The objective is to offer valuable insights for the research and development of crop breeding.

Training physical neural networks directly in matter remains difficult because most platforms do not implement weight storage and weight update within the same physical substrate. Here we show that engineered E. coli can implement a genetically encoded local learning rule acting on a persistent biological memory. In memregulons, analogue weights are stored as plasmid copy-number ratios in a coupled two-plasmid system and are rewritten by activity-dependent growth bias under a global negative learning signal. In single-strain cultures, theory predicts that the change in mean weight is proportional to the activity of the learning channel and to the standing variance of the stored distribution, and flow-cytometry trajectories across eight distinct promoters driving the learning channel support this prediction quantitatively. At the single-cell level, repeated negative learning also reshapes the stored distribution by narrowing it and increasing its skewness as weights approach the lower boundary. In mixed populations and nine-strain co-cultures, one global negative learning signal selectively rewrites only the active memregulons, enabling supervised adaptation in a bacteria-versus-bacteria tic-tac-toe tournament. We then generalise this principle across nine orthogonal chemical inputs and combinatorial promoters, including channels controlled by quorum-sensing molecules, and use it to rationally design a biological XOR gate. Finally, we examine multilayer ANN-like architectures with a human-in-the-loop protocol in which weight updates remain physically implemented and parameterised by experimental measurements, while inter-layer communication is supplied externally. These results establish a route to physical learning in living matter and provide a modular foundation for adaptive multicellular computation, paving the way for autonomous biological hardware capable of distributed environmental sensing and next-generation cellular therapeutics.

The ability to access, search, and analyse large collections of RNA molecules together with their secondary structure and evolutionary context is essential for comparative and phylogeny-driven studies. Although RNA secondary structure is known to be more conserved than primary sequence, no existing resource systematically associates individual RNA molecules with curated phylogenetic classifications. Here, we introduce PhyloRNA, a curated meta-database that provides large-scale access to RNA secondary structures collected from public resources or derived from experimentally resolved 3D structures. PhyloRNA allows users to search, select, and download extensive sets of RNA molecules in multiple textual formats, each entry being explicitly linked to phylogenetic annotations derived from five curated taxonomy systems. In addition to taxonomic information, each RNA molecule is accompanied by a rich set of descriptors, including pseudoknot order, genus, and three levels of structural abstraction - Core, Core Plus, and Shape - which facilitate comparative analyses across sets of molecules. PhyloRNA is publicly available at https://bdslab.unicam.it/phylorna/ and is regularly updated to incorporate newly available data and revised taxonomic annotations.

Protein function annotation is fundamental to understanding biological mechanisms, designing therapeutics, and advancing biomedical research. Current computational methods either rely on shallow sequence similarity or treat function prediction as isolated classification tasks, failing to capture the integrative reasoning across sequence, structure, domains, and interactions that expert biologists perform to infer function. We introduce BioReason-Pro, the first multimodal reasoning large language model (LLM) for protein function prediction that integrates protein embeddings with biological context to generate structured reasoning traces. A key input into BioReason-Pro is the set of GO term predictions made by GO-GPT, our autoregressive transformer that captures hierarchical and cross-aspect dependencies of GO terms. BioReason-Pro is trained via supervised fine-tuning on synthetic reasoning traces generated by GPT-5 for over 130K proteins and further optimized through reinforcement learning. It achieves 73.6% Fmax on GO term prediction and an LLM judge score of 8/10 on functional summaries, substantially outperforming previous methods. Evaluations with human protein experts show that BioReason-Pro annotations are preferred over ground truth UniProt annotations in 79% of cases. Remarkably, BioReason-Pro de novo predicted experimentally confirmed binding partners with per-residue attention localizing to the exact contact residues resolved in cryo-EM structures of those complexes. Together, GO-GPT and BioReason-Pro establish a framework for protein function prediction that combines precise ontology modeling with interpretable biological reasoning.

Plants have evolved sophisticated defense mechanisms to counter pathogen invasion, including the production of antimicrobial compounds, regulation of defense-related protein expression, and the synthesis of defense hormones across various subcellular organelles. While the significant contribution of organelle functions in plant immunity is increasingly recognized, the specific roles of these organelles in the immune response remain poorly understood. Recent studies have revealed that pathogen effectors from diverse microbes such as fungi, oomycetes, and bacteria localize within various organelles. These effectors target host proteins to manipulate the plant immune system, underscoring the crucial role of organelle functions in plant immunity. This review not only focuses on the localization of effectors within subcellular organelles, excluding the nucleus, but also explores the implications of organelle functions in the plant immune response. Gaining a deeper understanding of how these effectors interact with their targets in specific organelles will pave the way for developing disease-resistant plants.

Bioluminescent reporters are widely used to monitor and image biological processes. Among these, NanoLuc luciferase and its complementation variants (LgBiT/SmBiT and LgBiT/HiBiT) are commonly used due to their brightness, sensitivity, and compatibility with prolonged kinetic measurements. However, the single-channel emission of these NanoLuc-based systems (460 nm peak) limits their use in multiplexed assays. Prior efforts to shift NanoLuc’s emission employed bioluminescence resonance energy transfer (BRET) to a proximal fluorescent protein or organic fluorophore. Building on this concept, we engineered high-efficiency BRET reporters, termed NanoPrism luciferases, by inserting circularly permuted NanoLuc or LgBiT into a surface loop of the self-labeling HaloTag protein. These NanoPrisms achieve a ∼90% BRET efficiency by optimally positioning NanoLuc variants near a fluorophore covalently bound to HaloTag. The binary design further supports high- and low-affinity complementation, allowing applications in HiBiT knock-in cells and tracking protein–protein interactions, respectively. Pairing red-shifted NanoPrisms with unmodified NanoLuc or its complementation variants, we created a two-color bioluminescent reporter platform featuring bright signals of similar intensity and >100 nm spectral separation, allowing quantitative, simultaneous measurement of two molecular readouts within the same sample. Here, we demonstrate the platform’s utility for monitoring a degradation target alongside a control protein and for tracking two distinct events within a biological pathway, using plate-based detection and bioluminescence imaging. By enabling concurrent measurements within the same sample, the system provides insights into cellular dynamics while reducing variability and complexity associated with parallel single-channel assays.

The transfer of virulence factors into eukaryotic cells is a hallmark of bacterial pathogenesis. We report the expression, interspecies transfer, subcellular localization, and potential functions of three unusually large virulence factor-like proteins that underlie a bipartite mutualistic bacterial symbiosis. These proteins are synthesized by green sulfur bacterial epibionts surrounding a central motile chemoheterotroph in the multicellular phototrophic consortium 'Chlorochromatium aggregatum'. While symbiosis-proteins remain intracellular during axenic epibiont growth, they are transferred to the partner bacterium in the association. An RTX-like protein secreted towards the central bacterium is capable of degrading its alginate capsule, thereby promoting direct cell-to-cell contact. Two gigantic hemagglutinin-like proteins are predicted to fold when binding extracellular Ca2+ to form Type 6-like auto injection needles, explaining their observed transfer into the central bacterium. These functionalities extend far beyond the known pathogenic interactions of bacteria with eukaryotes and provide new perspectives on the evolution of bacterial virulence factors.

Plastic biodegradation in natural environments is increasingly recognized as a multi-organism process, yet the mechanisms enabling coordinated depolymerization and metabolism of polyethylene terephthalate (PET) remain poorly understood. Previously, we demonstrated that a full consortium containing three Pseudomonas and two Bacillus strains isolated from hydrocarbon-rich coastal soils of Galveston Bay, Texas, can synergistically depolymerize PET plastic and utilize it as a sole carbon source, a capacity not observed in individual isolates. In this report, using integrated comparative genomics, proteomics, and chemical analyses, we show that PET degradation in this system reflects exaptation of hydrocarbon metabolism reinforced by metabolic division of labor. Within this naturally occurring consortium, Bacillus strains persist under environmental stress, establish biofilms, and perform essential secondary hydrolysis, while Pseudomonas strains catabolize aromatic monomers and buffer oxidative stress. Genes supporting these functions are enriched within the accessory genomes of the consortium strains, indicating consortium-enriched horizontal gene transfer (HGT). In addition to the canonical two-step hydrolytic pathway well documented in PET biodegradation, we identify a secondary methylation- and redox-associated process, mechanisms where the full consortium acts on the oligomer mono(2-hydroxyethyl) terephthalate (MHET), yielding nearly complete conversion to terephthalic acid (TPA) and methylated MHET (MMHET). Together, these findings demonstrate how cooperation and competition within consortia facilitate targeted gene exchange, enabling emergent plastic biodegradation in natural microbial communities.

By mapping ribosome-protected fragments (RPFs) genome-wide, ribosome profiling (Ribo-seq) has uncovered extensive translation beyond conventional coding sequences, revealing non-canonical ORFs (ncORFs) with emerging roles in diverse biological processes. However, protocol-induced biases introduced during library construction can substantially distort RPF signals. Most existing ORF callers are not designed to explicitly account for such artifacts, limiting robust ncORF identification. Here, we present RiboBA, a bias-aware probabilistic framework to address this challenge. RiboBA consists of two main components: a generative module that recovers protocol-induced biases and codon-level ribosome occupancy, and a supervised module that identifies translated ORFs and initiation sites using the resulting bias-adjusted profiles. Evaluated through simulations and on a range of Ribo-seq datasets-particularly supported by cell-type-specific immunopeptidomics-RiboBA robustly recovers protocol-induced parameters and achieves superior accuracy and sensitivity in ncORF identification. Notably, RiboBA performs particularly well on RNase I libraries with attenuated three-nucleotide periodicity, as well as on MNase and nuclease P1 libraries, while maintaining competitive runtimes. In a Drosophila case study, RiboBA identifies conserved ncORFs with coding potential, including recurrent upstream translation of ThrRS and Mettl2 that suggests a potential threonine-specific translational control axis.