Ammonia oxidation is a rate-limiting step in the nitrogen cycle, yet viral contributions to this process remain largely unresolved. Here, we identify three genomically distinct groups of amoC-encoding phages (155-338 kilobases in length; termed as amoC-phages) from multiple freshwater lakes in Europe and North America, including the Laurentian Great Lakes. These phages are highly divergent in phylogeny, genome architecture, and gene content, and are predicted to infect two distinct Nitrosomonadaceae ammonia-oxidizing bacterial lineages. The placement of phage-encoded amoC genes across these divergent viral clades indicates independent acquisition of amoC. Time-series and depth-resolved metagenomes and metatranscriptomes reveal persistent and depth-structured distributions of amoC-phages and their predicted hosts, with seasonal mixing periodically reshaping their co-occurrence patterns. Furthermore, virome data from Lake Mendota show that some of the amoC-phages occur as free viral particles, supporting active viral lysis and particle redistribution along the water column. Metatranscriptomes of the Laurentian Great Lakes reveal coordinated expression of phage structural genes (e.g., major capsid protein) together with phage-encoded amoC, indicating active infection in situ. Together, these results support a framework in which amoC-phage infection is depth-structured, seasonally dynamic, and coupled to ammonia-oxidizing bacterial host activity, highlighting viruses as previously overlooked components of freshwater nitrogen cycling.

Serine integrases enable precise genome engineering but remain underused for iterative genome integration because current workflows are often labor-intensive and difficult to scale. Here, we report SARGE, a serine integrase-assisted rapid genome integration platform, that enables iterative insertion of large DNA fragments into the E. coli genome through cassette exchange. By combining the orthogonal serine integrases PhiC31 and Bxb1 with rational donor plasmid design, SARGE supports rapid, programmable integration cycles without the need for resistance marker excision between rounds. To simplify the identification of correct recombinants, we incorporated a dual-fluorescence reporter system based on sfGFP and mKate and developed a green–red screening strategy for direct, naked-eye colony selection. This visual screen identified the desired recombinants with 100% accuracy among the colonies tested. SARGE achieved cassette exchange efficiencies of up to 95% and maintained efficiencies of approximately 90% for cargoes as large as 10 kb. Together, these features substantially streamline iterative genome integration and establish SARGE as a robust and accessible platform for genome engineering and synthetic biology in E. coli, with potential for extension to other genetically tractable microbial hosts.

RNAs adopt complex structures that regulate key biological processes, making accurate structure prediction essential. Chemical probing coupled with Nanopore direct RNA sequencing (DRS) offers a route to single-molecule structural inference, but current tools are limited by inaccurate signal-to-sequence alignment, which degrades modification-rate estimation and downstream structure prediction. Here we introduce segSHAPE for RNA secondary structure prediction from Nanopore DRS data (both RNA002 and RNA004 chemistries), a probe-agnostic framework that improves signal alignment using prior information of basecalling and per-read signal baseline shift correction, learns position-specific k-mer raw signal parameters, and estimates per-nucleotide modification rates with an unsupervised anomaly detector. On three public RNA002 DRS datasets spanning different chemical probes (AcIm, NAI-N3) and RNAs from 421 to 1552 nt, segSHAPE achieves the highest F1 score and Matthews correlation coefficient (MCC) on all RNAs, exceeding the strongest baseline by 3.4 to 5.8 percentage points in MCC. It additionally captures the ligand-induced conformational change of the thiamine pyrophosphate (TPP) riboswitch RNA directly from RNA002 DRS data using the DEPC probe. On a public RNA004 DRS dataset, segSHAPE improves over the sm-PORE-cupine baseline by 17 ROC-AUC points in modification rate estimation and by 6.7 MCC points in structure prediction. These results establish segSHAPE as a unified, probe-agnostic pipeline for RNA structure prediction from Nanopore DRS data.

Generative genome design models can now produce previously unobserved genome-length sequences, but assessing their capabilities is complicated by limitations in functional prediction. The ability to engineer genomes faster than we can understand them risks creating biosecurity vulnerabilities. To evaluate these potential risks systematically, we propose a framework that distinguishes between (i) evolutionary novelty, quantified through phylogenetic and sequence similarity to natural genomes; and (ii) design efficiency - the efficiency with which a model finds viable sequences compared to simple baseline generators. Applying this framework to bacteriophages designed by the genome language model Evo 2, we find that model likelihood strongly predicts experimental viability, capturing functional constraints beyond simple biological heuristics. However, this efficiency derives largely from staying close to previously observed sequences rather than exploring novel sequence space, reflecting the combined performance of the model and additional filters that were applied to its outputs. Compared to baselines of random mutagenesis and serial passage, the model achieves substantial design efficiency while its outputs remain phylogenetically close to natural genomes. We conclude that the generative capabilities of Evo 2 warrant low to moderate biosecurity concern for de novo hazard creation, although the degree to which these findings generalise to larger or less constrained viral architectures is an open question. Our framework enables an evidence-based capability assessment of generative genome design tools, informing future biosecurity evaluations.