Depth-governed ecological and evolutionary partitioning of ocean trench viromes

Depth-governed ecological and evolutionary partitioning of ocean trench viromes | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Depth-governed ecological and evolutionary partitioning of ocean trench viromes Karthik Anantharaman, Yuan-Guo Xie, Jianxing Sun, Yanling Qi, and 11 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8682536/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Viruses influence microbial mortality and carbon flux in the ocean, yet how water-column depth shapes their ecology and evolution from surface waters to the hadal zone remains unexplored. To address this gap, we profiled a continuous 2–6,000 m transect in the Yap Trench in the Pacific Ocean, amongst the world’s deepest ocean trenches using depth-resolved metagenomics. We reconstructed a total of 8,520 viral operational taxonomic units and observed the viral communities to be strongly stratified, with diversity peaking in the mesopelagic. Virulent Kyanoviridae dominated upper layers, while temperate Peduoviridae increased and stabilized with depth, indicating a shift from "kill the winner" to "piggyback the winner" dynamics. Auxiliary metabolic genes mirrored this partitioning but were less defining than lifestyle transitions. Viruses infecting Gammaproteobacteria and Firmicutes encoded cell motility and stress-response genes, suggesting adaptation to energy limitation. Evolutionary dynamics were depth-dependent: mutation rates declined with depth, while positive selection intensified in key viral genes. Overall, our results identify ocean depth as a central axis shaping viral community dynamics and evolutionary trajectories in marine ecosystems. Biological sciences/Ecology/Biooceanography/Microbial biooceanography Earth and environmental sciences/Ecology/Microbial ecology Biological sciences/Microbiology/Bacteriophages Biological sciences/Microbiology/Environmental microbiology/Water microbiology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Main Viruses are the most abundant biological entities in the ocean and exert profound influence over microbial dynamics and marine biogeochemical cycles 1 , 2 . By inducing host lysis, they release large pools of organic and inorganic nutrients, fueling the microbial loop and sustaining trophic connectivity across pelagic food webs 3 , 4 . Beyond their role in biomass turnover, viruses also shape microbial evolution via horizontal gene transfer, including the delivery of auxiliary metabolic genes (AMGs) that reprogram host metabolism and mediate key geochemical transformations 5 , 6 . These dual roles, regulating microbial population dynamics and modulating ecosystem function, position viruses as key ecological agents in global biogeochemical cycling 7 , 8 . Yet despite this ecological importance, much of our understanding of marine viruses stems from studies conducted in surface or euphotic waters 9 , 10 , 11 , 12 . How viral communities adapt and function in deeper oceanic layers, especially in the hadal zone, remains poorly resolved, largely due to sampling and technical challenges. Oceanic trench represents one of the most isolated and extreme habitats on Earth, formed at tectonic subduction zones and descending to depths beyond 6,000 mbsl 13 , 14 . These environments are characterized by intense hydrostatic pressure (> 1,000 atm), near-freezing temperatures, complete darkness, and severe nutrient limitation 15 . Despite these constraints, trench ecosystems harbor diverse and metabolically specialized microbial life 16 , 17 . Recent metagenomic studies have uncovered rich microbial and viral diversity in trench sediments 18 , 19 , 20 , including novel AMGs linked to deep biosphere functions 21 , 22 , 23 . Yet, viral communities in the overlying water column, especially transitions from epipelagic to mesopelagic and further to hadal zones, remain largely unexplored. This gap hinders our understanding of how viral diversity, function, and evolution are structured by depth-dependent environmental pressures. To address these gaps, we conducted depth-resolved metagenomic profiling of viral communities across the full vertical profile of the Yap Trench (137.52–138.70 °E, 8.02–9.90°N), spanning from surface (2 mbsl) to hadal depths (> 6,000 mbsl). Here, using depth-resolved metagenomics of 17 water samples spanning the entire Yap Trench water column, we investigated how environmental gradient’s structure viral community composition, mediate host interactions, and further shape unveil evolutionary trajectories across ocean depth. Our findings together establish a depth-stratified framework for understanding viral persistence, adaptation, and biogeochemical influence across Earth’s deepest marine environments. Results Viral community profiles and physicochemical characteristics of oceanic trench viromes Following sample collection from the Yap Trench, physicochemical analyses revealed pronounced vertical gradients associated with increasing depth (Fig. S1 b). Temperature declined sharply from surface levels to the hadal zone, dropping from near 29°C to approximately 1.5°C. Dissolved oxygen followed a non-linear distribution, initially decreasing through the mesopelagic zone before gradually increasing in deeper layers. Dissolved organic carbon (DOC) exhibited a mid-depth maximum between 200 and 300 mbsl, while both surface and hadal layers maintained relatively low DOC concentrations. In contrast, nitrate exhibited elevated concentrations in the deep zones. Additional physicochemical parameters are detailed in Supplementary Data 1. Such depth-stratified environmental gradients are likely key drivers of viral community dynamics, particularly across the mesopelagic-to-hadal transition zone. To investigate depth-associated changes in viral community, we conducted a comprehensive analysis of 17 metagenomic datasets spanning the Yap Trench water column. A total of 11,184 viral contigs with length ≥ 10 Kbp were recovered, forming the basis for subsequent profiling. Clustering these sequences at a 95% average nucleotide identity threshold yielded 8,520 viral operational taxonomic units (vOTUs), serving as a comprehensive representation of the viral community structure throughout the water column (Fig. S2). Yet, only 3.6% and 9.9% of these vOTUs matched entries in the Global Ocean Virome database v2 and IMG/VR v4 databases, respectively (Fig. S3), underscoring the exceptional genomic novelty of the Yap Trench virome. Among them, 188 and 499 vOTUs were categorized as high-quality and medium-quality, respectively, based on completeness ≥ 90 and ≥ 50 (Supplementary Data 2). Sharp decline in viral diversity below 500 mbsl marks a pivotal ecological boundary Viral diversity, in terms of both richness and abundance, exhibited consistent vertical trends throughout the water column, peaking before a sharp decline in the mesopelagic zone (100–300 mbsl), highlighting the pronounced vertical stratification of viral communities along the depth gradient (Fig. 1 a and 1 b). This vertical structuring was further supported by the distribution of vOTUs, the majority of which showed strong sample specificity (Fig. S4). Only a limited fraction of vOTUs was shared across multiple samples, underscoring the pronounced spatial heterogeneity and vertical differentiation of viral assemblages in the deep-sea water column. Of the 8,520 vOTUs identified, 5,742 (67.4%) could be classified into 99 viral families (Supplementary Data 2). Surface and mesopelagic communities predominated by Kyanoviridae (average 53.64%) and Autographiviridae (average 9.46%), while deep-water assemblages (500–6000 mbsl) were enriched in unclassified taxa (average 48.97%), alongside Hendrixvirinae (average 3.71%) and Peduoviridae (average 12.93%). The sharp decline in viral abundance between mesopelagic and deeper waters was largely attributable to the marked reduction of Kyanoviridae in deep layers. Principal coordinate analysis (PCoA) revealed three distinct clusters of viral communities stratified by depth: 2–50 mbsl for surface zones, 100–300 mbsl for mesopelagic zones, and 500–6000 mbsl for deep zones, further corroborating the pronounced vertical structuring of these communities (Fig. 1 c and d; R 2 = 0.55, p < 0.001). Even without the consideration of Kyanoviridae , the viral communities still clustered into the same three groups (Fig. S5), suggesting both abundant and rare viruses affect the viral community composition across different layers. Canonical correspondence analysis (CCA) identified depth and salinity as the principal environmental axis structuring viral communities. These two factors show an approximately orthogonal relationship, with depth governing viral assemblages in deep layers and salinity shaping those in surface layers. In contrast, DOC was more closely aligned with mesopelagic communities, suggesting distinct resource-driven dynamics at intermediate depths (Fig. 1 d). These relationships between environmental factors and viral communities mirrored, to a considerable extent, those observed for microbial communities (Fig. S6). Together, these data identify the mesopelagic-to-bathypelagic transition as a critical ecological inflection point in viral community structure, likely driven by a combination of abiotic and biotic factors. Host specificity and viral lifestyle underlie depth-specific community transitions To uncover the biological mechanisms underlying these community shifts, we next examined viral lifestyles and host associations, aiming to identify how host biogeography and infection strategies influence viral persistence across depths. We identified 697 (7.89%) vOTUs as temperate viruses and 3,525 (43.43%) vOTUs as virulent viruses (Fig. 1 e; Supplementary Data 2). The richness of both temperate and virulent viruses peaked in the mesopelagic zones but declined sharply beyond the mesopelagic layers, with virulent viruses consistently maintaining overwhelming diversity across all depths (Fig. 1 b). Notably, the abundance of temperate viruses remained comparatively stable throughout the water column, whereas virulent viruses showed pronounced variability (Fig. 1 f). Their populations were elevated in the surface and mesopelagic zones, but declined sharply at greater depths. These patterns indicate that vertical changes in viral diversity are primarily driven by fluctuations in virulent viral populations. Such depth-dependent shifts in the balance between virulent and temperate viruses imply a transition in viral–host strategies: lytic infections dominate in the shallower layers, consistent with kill-the-winner dynamics, whereas the deeper layers show a modest but detectable rise in temperate lifestyles, signaling a shift toward a piggyback-the-winner mode as host turnover declines (Fig. 1 f). By applying a computational host-prediction framework (see Methods), we assigned 1,760 vOTUs to putative microbial hosts (Supplementary Data 2). Viruses linked to Gammaproteobacteria , Alphaproteobacteria , Bacteroidota , and Chloroflexota predominated in surface waters, while archaeal viruses, particularly those infecting Nitrososphaeria and Poseidoniia , were largely confined to mesopelagic depths, in line with host distributions (Fig. 2 a–c). An exception was observed for the Cyanobacteria-Cyanophage pair: Cyanobacteria were most abundant in surface waters, but cyanophages peaked in the mesopelagic zone, indicating spatial decoupling between host and virus. Viral contigs from deep waters exhibited a higher rate of host assignment compared with surface and mesopelagic layers. This is likely caused by the simplified community structure dominated by Gammaproteobacteria and Firmicutes . Among these, viruses infecting Gammaproteobacteria were consistently abundant across all depths, presumably due to the broad ecological range of their hosts. For instance, some lineages such as Alteromonadaceae are ubiquitous throughout the water column, whereas others exhibit marked depth specificity (Fig. S7). It is noteworthy that most Gammaproteobacteria-infecting viruses were affiliated with Kyanoviridae and Peduoviridae , representing the most abundant virulent and temperate lineages, respectively (Fig. 2 d). Both lineages also infect additional microbial groups, albeit with contrasting host ranges. Kyanoviridae display a broad host spectrum, extending to major surface and mesopelagic taxa such as Bacteroidota and Cyanobacteria . These host taxa peaked in the mesopelagic zone but declined sharply below 300 mbsl (Fig. 2 c). In contrast, Peduoviridae exhibit a narrower range, primarily targeting Gammaproteobacteria and Firmicutes , two lineages that are particularly enriched in deep-sea environments. Together, these findings suggest that host biogeography and lifestyle strategy are key factors driving viral community turnover with depth. Viruses reprogram host metabolism via depth-stratified AMGs To determine whether viruses influence host function in a depth-dependent manner, we examined the distribution and taxonomic context of AMGs across the water column. Virally encoded AMGs span diverse host pathways involved in key biogeochemical cycles (Fig. 3 a), highlighting their roles in shaping microbial function and ecosystem processes. Viral AMGs displayed clear vertical stratification in both function and host association, reflecting depth-specific metabolic adaptation. In the mesopelagic zone (100–300 mbsl), viruses exhibited a high diversity of AMGs involved in key metabolic functions such as carbohydrate utilization ( galE , ppdK , glgM ) and nitrogen and sulfur cycling ( norQ , nosD , cycC , cycH , Fig. 3 b, Supplementary Data 3). The observed functional repertoire may in part be attributable to the richness of viral taxa in mid-water environments. These AMGs were primarily encoded by viruses infecting Bacteroidota , Chloroflexota , and Cyanobacteria , taxa known to be metabolically active at mid-depths. Notably, photosynthesis-related AMGs (e.g., psbA and petE ) also peaked in the mesopelagic and were broadly associated with diverse microbial taxa. Phylogenetic placement showed that psbA genes clustered with cyanophage homologs (Fig. S8), suggesting lateral gene transfer and broader dissemination beyond canonical cyanophages. This implies that viruses may help maintain host phototrophic near the lower edge of the photic zone, potentially extending photosynthetic functionality under low-light conditions. Additionally, AMGs related to nucleotide metabolism and cofactor biosynthesis were most enriched in mesopelagic zones, potentially facilitating host biosynthetic processes and promoting viral proliferation under moderately favorable energetic conditions. In contrast, deep-sea viral communities (> 500 mbsl) exhibited a marked depletion of nutrient cycling AMGs (Fig. 3 b), potentially resulting from the limited viral–host interactions characteristic of deep-sea environments. However, genes related to environmental sensing and cellular adaptation, such as signal transduction, transport, and stress response, were detected in deep-sea viral communities (Fig. 3 b). These included two-component regulators ( csrA , yesN , narL ) and motility-related genes ( pilT , cpaB ), largely encoded by viruses infecting Gammaproteobacteria and Firmicutes . Such functions may enable hosts to perceive physicochemical stressors and optimize the allocation of scarce resources, enhancing fitness in energy-limited deep-sea environments. These findings suggest that viruses infecting Gammaproteobacteria and Firmicutes may influence host metabolism and motility, potentially enhancing host fitness under deep-sea conditions. Collectively, these data reveal that viral AMGs are reflect a layered functional architecture that aligns with host ecology and environmental context. Viral reprogramming of host metabolism is most pronounced in the mesopelagic zone, where energy and nutrient gradients intersect, while deeper waters favor adaptation-related functions. This stratified pattern suggests that viruses play depth-specific roles in modulating host ecological performance across the oceanic water column. Depth imposes selective pressures shaping viral microevolution Given the strong functional stratification of viral genes across depth layers, we further investigated whether such adaptations are evolutionarily conserved and under selective pressure. Specifically, we examined how mutation rates and selection regimes vary across ocean depths to reveal the microevolutionary processes underpinning viral persistence in distinct environments. With increasing depth, the marine environment becomes progressively more extreme, characterized by persistent low temperatures, high hydrostatic pressure, and severe energy and nutrient limitation 24 , 25 . These physicochemical constraints likely impose intensified selective pressures, thereby contributing to the marked decline in viral diversity observed in deeper waters. To evaluate this hypothesis, we further examined the microevolutionary trajectories of viruses across different depth layers. Our results showed that the proportion of vOTUs carrying mutations was lower in the deep layer (Fig. 4 b; Supplementary Data 4), but the diversity of mutated nucleotides per loci tended to be higher (Fig. S9). Mutation density (SNVs per Kbp) also declined markedly with depth, with median values of 19.4 at the surface, 15.8 in the mesopelagic, and 12.0 in the deep ocean (Fig. 4 a). This suggests the existence of different evolutionary trajectories of viruses across depth layers. Surface and mesopelagic viruses exhibited a larger pool of mutations, with more genomic sites affected. In contrast, deep-sea viruses accumulated fewer mutations overall, but individual sites displayed greater nucleotide diversity. This pattern may increase the likelihood of beneficial variants being preserved under specific evolutionary circumstances. Consistent with this, viruses displayed progressively relaxed purifying selection with depth, as indicated by increasing median pN/pS values, from 0.106 at the surface to 0.141 in the mesopelagic and 0.148 in deep waters (Fig. 4 e). Moreover, the proportion of genes under positive selection was significantly higher in deep-sea viruses compared with those from the upper layers (Fig. 4 f). These trends remained robust when viruses were further divided by lifestyle into virulent and temperate groups (Fig. 4 c, d, g and h). Taxonomically, these genes were associated with dominant microbial host lineages at each depth (Fig. S11a). In absolute numbers, however, virulent viruses peaked in the mesopelagic zone, whereas temperate viruses increased steadily with depth (Fig. S11b). By examining individual viruses, we found that higher SNV counts were mostly associated with dominant viruses. Those infecting Alphaproteobacteria , Gammaproteobacteria , Chloroflexota , Cyanobacteria , and Bacteroidota predominated in the surface and mesopelagic zones, whereas viruses targeting Gammaproteobacteria and Firmicutes prevailed in deeper waters (Fig. S10a). Positively selected genes were also more frequently observed in these abundant viruses (Fig S11a), suggesting that positive selection strengthens their competitive capacity and further promotes their fitness and proliferation. This trend was particularly pronounced in deeper layers, where the proportion of viruses under positive selection increased with relative abundance—from ~ 20% in surface waters to ~ 50% in the mesopelagic and ~ 70% in the deep ocean (Fig. 5 a). Functional annotation of these loci revealed depth-specific divergence shaped by environmental selection, though only 29.1% could be assigned putative functions (Fig. S12; Supplementary Data 5). Across all depths, structural and host-interaction genes consistently dominated, suggesting strong selection on viral entry, recognition, and infection, consistent with an ongoing virus–host arms-race dynamics (Fig. 5 b). Lifestyle comparisons further showed that temperate viruses were enriched in genes related to genetic information processing and metabolism, whereas virulent viruses predominantly exhibited selection on host-recognition and structural assembly (Fig. S13), pointing to contrasting adaptive strategies between lifestyles. We next examined positively selected AMGs to resolve how metabolic potential contributes to viral adaptation. The 2OG–Fe(II) oxygenase gene was the most consistently enriched AMG across depths, peaking in the mesopelagic zone and largely encoded by viruses infecting depth-specific dominant microbial lineages (Fig. 5 c, Fig. S14). Its enrichment likely reflects selective pressures favoring hydroxylation-mediated pathways that support host metabolism under nutrient- and oxygen-limited conditions 26 , 27 . Beyond this, functional divergence of AMGs was apparent across depths: nitrogen- and sulfur-cycling genes were enriched in the mesopelagic zone, suggesting viral enhancement of host biogeochemical roles, whereas deep-sea AMGs were dominated by nucleotide metabolism, such as cytosine methylation (DNA (cytosine-5)-methyltransferase, dcm ) encoded largely by Gammaproteobacteria-infecting viruses (Fig. 5 d). Noticeably, most positively selected AMGs were encoded by virulent viruses across all depths (Fig. S15), whereas temperate virus-associated AMGs were largely restricted to the deep ocean and primarily linked to cofactor biosynthesis and nucleotide metabolism. Collectively, these results suggest that positively selected AMGs follow depth-stratified, lineage-specific, and lifestyle-dependent patterns, highlighting their potential role in shaping host adaptation and ecosystem functioning across the oceanic water column. Discussion Viruses exert strong control over marine microbial processes, yet their ecological and evolutionary dynamics across depth remain poorly resolved due to deep-sea sampling challenges. Previous studies have largely focused on surface and midwater communities, leaving the deep ocean, particularly hadal zones, underexplored. By establishing a full-depth continuum from the epipelagic to the hadal zone in Yap Trench and applying high-resolution metagenomics and advanced bioinformatics, our study overcomes these constraints and provides a more holistic view of how depth shapes viral diversity, virus–host interactions, and adaptive evolution. Viral diversity peaked in the mesopelagic zone, where DOC had the most prominent effect on their distribution. This could be attributed to the favorable conditions for virus-host interactions, characterized by moderate nutrient availability, reduced UV exposure, and a permissive thermal regime 28 , 29 , 30 . The dominance of virulent over temperate viruses suggests a kill-the-winner dynamic and an efficient viral shunt, with continuous lytic turnover injecting bioavailable DOC 2 , 31 , 32 that fuels heterotrophic respiration and elevates oxygen demand. Although direct causality cannot be established, the concordance between DOC maxima, lytic dominance and steep O 2 decline beneath 200–300 m suggests that viral lysis contributes to shaping the upper margin of the OMZ. This DOC–lysis coupling provides a plausible mechanism linking mesopelagic DOC enrichment and viral activity to enhanced carbon recycling and oxygen drawdown in the ocean interior (Fig. 6 ). Field observations reinforce this interpretation, showing positive correlations between modeled carbon flux and viruses infecting phytoplankton and picoplankton, and demonstrating virus-induced aggregation and export of phytoplankton biomass into mesopelagic depths 33 , 34 , 35 . This midwater peak contrasts with deeper layers, where viral diversity declines sharply. The reduction likely stems from limited flux of labile organic matter, as sinking particles take weeks to months to reach hadal depths and arrive largely refractory, restricting microbial and viral activity 36 , 37 . Harsh physicochemical conditions—high pressure, low temperatures, and energy scarcity—further constrain replication 38 . Under such conditions, infrequent host encounters make lytic infection costly, favoring temperate strategies 42 , 43 . Virulent virus abundance thus decreases with depth, though their richness remains relatively high, suggesting many lineages persist in dormant or low-activity states with potential for reactivation 39 , 40 . Field observations support this hypothesis, showing detectable but reduced lytic activity in bathypelagic waters and a gradual virus-to-prokaryote ratio decline 41 . By contrast, temperate viruses decline more slowly and can become relatively enriched in deep communities 42 , reflecting a depth-driven shift from lytic dominance to lysogenic persistence. A major contributor to this vertical contrast is the family Kyanoviridae . Dominant in surface and mesopelagic waters but nearly absent in the deep ocean, Kyanoviridae largely account for the difference in community composition between midwater and hadal layers. Although typically recognized as cyanophages infecting Prochlorococcus and Synechococcus 43 , their unexpectedly high abundance in the mesopelagic layer (200–300 m), where phototrophic cyanobacteria are relatively scarce, can be explained by multiple processes. A legacy effect may contribute, with large numbers of viral particles produced in surface waters transported downward through water column mixing and particle flux. The mesopelagic zone also often coincides with the deep chlorophyll maximum or the upper boundary of the oxygen minimum zone, where residual cyanobacterial activity continues to support viral replication 44 . In addition, some Kyanoviridae may possess broader host ranges, potentially infecting additional bacterial lineages beyond cyanobacteria (e.g. Pseudomonadota , and Actinomycetota ) 45 , 46 , 47 . Enhanced microbial respiration and organic matter availability in this layer foster a metabolically active microbial community, which indirectly sustains Kyanoviridae persistence even where phototrophic cyanobacteria are scarce, thereby explaining their high abundance in the mesopelagic zone compared with the deep ocean. Below the mesopelagic, viruses infecting Gammaproteobacteria become increasingly dominant, reaching their highest prevalence in the deep ocean, in contrast to the cyanophage-dominated upper layers. This vertical structuring reflects both host turnover and shifts in infection strategies. In nutrient-rich surface and mesopelagic waters, Kyanoviridae and Autographiviridae prevail, with lytic infections fueling rapid host turnover and organic matter recycling, consistent with the dynamic nature of upper-ocean ecosystems 47 , 48 . These viruses are chiefly associated with the D2472 and TMED112 families, SAR86-lineage Gammaproteobacteria characterized by extreme genomic streamlining, proteorhodopsin-enabled photoheterotrophy, and enhanced DOM-scavenging capacities, well adapted to upper-ocean environments 49 , 50 . At greater depths, however, temperate Peduoviridae become prominent (Fig. S6a), echoing reports of lysogeny as a hallmark of energy-limited deep-sea communities 21 . Their hosts, mainly including Alcanivoracaceae and Oleiphilaceae , are lineages specialized for survival under energy-limited deep-ocean conditions 51 , where lysogeny provides a selective advantage against resource scarcity. Thus, depth-related transitions in Gammaproteobacteria-virus interactions encapsulate the broader ecological restructuring that shapes the microbial life throughout the water column. Our results support a unified “dual-filter” framework in which depth governs both ecological opportunities and evolutionary trajectories of marine viruses, linking environmental gradients to viral adaptation. Depth-stratified viruses follow distinct evolutionary trajectories. Notably, elevated mutation rates in surface and mesopelagic layers did not translate into higher pN/pS values, indicating limited adaptive substitutions. This paradox can be explained by mutation–selection dynamics: while increased mutation rates raise the chance of beneficial variants; excessive mutational input generates a heavy deleterious load that erodes the fitness advantage of adaptive alleles 52 , 53 , 54 . Stronger purifying selection in surface waters further supports this view, suggesting that most mutations were deleterious. As a result, natural selection becomes less efficient, and the proportion of positively selected substitutions remains low despite abundant mutational supply. Evidence points to ecological variability as a potential driver of viral microevolution 55 . Compared to surface and mesopelagic layers, deep layers are more extreme with reduced light, lower temperature, higher pressure and limited nutrients 56 , 57 , 58 . These conditions favor a shift in viral lifestyles from virulent to temperate, allowing better adaptation to adverse environments 59 . Temperate viruses can accelerate adaptive change via long-term host association, frequent gene acquisition and recombination, and carriage of accessory genes that are exposed to selection, which has been observed experimentally and in large-scale genomic surveys 60 . Strikingly, the most direct evidence for this process in the deep ocean is the pervasive positive selection on integrase genes (Fig. 5 d). Such selection likely enhances integration efficiency, thereby stabilizing lysogeny and boosting host fitness under conditions of low cell density and chronic energy limitation 61 , 62 , 63 . This evolutionary signal underscores the pivotal role of temperate strategies in driving horizontal gene transfer, positioning lysogeny as a major force in restructuring host genomes and promoting long-term adaptability in deep-sea ecosystems 64 , 65 . Moreover, adsorption- and entry-related structural genes (for example, tail and tail-fiber loci) also experienced positive selection (Fig. 5 d). Even in lysogeny-dominated deep-sea communities, temperate phages must occasionally re-enter transmission phases where adsorption efficiency and host range directly determine fitness 42 . Tail/tail-fiber receptor-binding proteins, particularly at ligand-contact tips, thus exhibit episodic diversifying selection. Mechanistically, this is expected, since these modules operate at the frontline of host recognition and immune evasion 66 , 67 , 68 . These dynamics suggest that even in energy-limited deep waters, sporadic but critical host encounters within microhabitats sustain antagonistic coevolution, sustaining adaptive evolution in entry modules alongside the dominance of lysogeny. Materials and Methods Sampling, physicochemical assay, DNA extraction and metagenomic sequencing Seawater samples were collected from seventeen depths ranging from 2 mbsl to 6,000 mbsl above the Yap Trench during research cruises in June 2017. At each depth, approximately 10 L of seawater was filtered through 0.22 µm pore-size membrane filters (Millipore, MA, USA) to capture microbial biomass. Filters were immediately flash-frozen in liquid nitrogen and stored at − 80°C until further processing. Environmental parameters, including temperature, salinity, dissolved oxygen, and nutrient concentrations, were measured following established protocols 69 and are summarized in Supplementary Data 1. Total DNA was extracted from filters using the PowerSoil DNA Isolation Kit (MoBio Laboratories, Carlsbad, CA, USA), following the manufacturer’s instructions. Metagenomic libraries were prepared and sequenced on an Illumina MiSeq platform using 2 × 150 bp paired-end reads, generating an average of ~ 38 Gbp of raw data per sample. Two additional metagenomic datasets generated from the same project were integrated for further analyses 70 . Assembly and binning of metagenomics Metagenomic raw reads were quality filtered using fastp (v0.23.2) with parameters “q 20 -u 20 -l 50” to eliminate adapter sequence and low-quality bases 71 . Filtered reads were assembled using SPAdes (v3.15.2) with the parameters “-k 21, 33, 55, 77, 99,127 -meta” 72 . Scaffolds from each assembled sample were aligned against the cleaned reads from all depth layers using BBMap (v38.92; http://sourceforge.net/projects/bbmap/ ) with parameters “k = 15 minid = 0.9 build = 1”. Scaffold coverage across samples was calculated using the jgi_summarize_bam_contig_depths script in MetaBAT (v2.12.1). Genome binning for each sample was performed on scaffolds using MetaBAT (v2.12.1), based on differential coverage information across depth layers 73 . The completeness and contamination of MAGs were assessed using CheckM (v1.1.3) 74 . The taxonomy of MAGs was obtained using GTDB-tk (v2.3.0) with reference to the genome taxonomy database (GTDB; release 214) 75 , 76 . The universal prokaryotic marker gene Ribosomal Protein S3 ( rpS3 ) was identified following established methods 77 . Identification of viral genomes Putative viral contigs were identified using VIBRANT (v1.2.1) and VirSorter2 (v2.2.3), both run with default parameters 78 , 79 . For VirSorter2, only contigs with max_score ≥ 0.90 were retained as high-confidence viral sequences. In accordance with the minimum information about an uncultivated virus genome (MIUViG) standard 80 , viral contigs ≤ 10 kb in length were excluded from downstream analyses. CheckV (v0.8.1) was used to assess the completeness and contamination of viral genomes, and those lacking viral genes were subsequently removed 81 . vOTUs were defined by clustering contigs at 95% nucleotide identity across 85% alignment fraction using CD-HIT-EST (v4.8.1) with parameters -c 0.95 -aS 0.85 82 . To assess the similarity between Yap Trench viruses and previously characterized marine viromes, we retrieved reference datasets from the IMG/VR v4 83 and Global Ocean Viromes 2.0 (GOV2) databases 84 . Viral lifestyle predictions were performed using DeePhage (v1.0), with probability scores ≤ 0.20 interpreted as temperate and ≥ 0.80 as virulent lifestyles 85 . Taxonomic classification of viral genomes was performed using PhaGCN2.0 86 , a deep learning-based tool for virus taxonomic assignment. Open reading frames (ORFs) were predicted using Prodigal (v2.6.3) in metagenomic mode (-p meta) 87 . Predicted proteins were annotated against multiple databases, including NCBI-nr, KEGG and EggNOG v5.0 using DIAMOND (v2.0.14.152) with an E-value threshold of 1e − 5 88, 89 90 . Furthermore, DRAM-v.py (v1.5.0) was employed for functional annotation to refine the understanding of viral-encoded metabolic functions 91 . For viral genes under positive selection but lacking functional annotations, protein structures were predicted using ColabFold (v1.5.2) 92 , and subsequent structural comparisons were performed with Foldseek (10-941cd33) to infer potential biological roles 93 . Viral host prediction was conducted using the iPHoP pipeline with default parameter 94 . To improve prediction resolution, the reference database was extended to include 734 MAGs from this study in combination with the native iPHoP database. Calculating microbial and viral abundance The abundance of vOTUs and rpS3 genes was quantified using average number of aligned reads overlapping each position (trimmed_mean) metric. Clean reads were mapped to datasets using CoverM (v0.6.1) with “make” parameters to generate alignment BAM files 95 . Reads not meeting alignment criteria were filtered using “filter” module with the following parameters: “--min-read-percent-identity 95 --min-read-aligned-percent 75”. Final abundance values were calculated using the contig module with trimming and alignment thresholds set as “--trim-min 0.10 --trim-max 0.90 --min-read-percent-identity 0.95 --min-read-aligned-percent 0.75”. The abundance of each vOTUs or gene was estimated as the average per-base read coverage, normalized by the total read count of the corresponding library, and scaled to the mean read count across all 17 libraries. Analysis of viral populations Clean reads from each sample were mapped to the concatenated set of all representative vOTUs using Bowtie2 with default parameters 96 . Single nucleotide variants were identified using inStrain (v1.5.4) in --database-mode with default parameters 97 . SNV calling required a minimum site coverage of 5×, and only variants with a minor allele frequency ≥ 5% were retained. To reduce false positives, variant alleles were required to exceed the expected Illumina sequencing error rate (1 × 10⁻⁶). Genome-wide nucleotide diversity (SNVs per kilobase) and the ratio of nonsynonymous to synonymous substitutions (pN/pS) were calculated at both the genome and gene levels. The pN/pS ratio was used as a proxy for selective pressure, where values > 1 indicate potential positive selection, and values < 1 reflect purifying selection. Statistical analysis All statistical analyses were conducted in R (v4.1.3). Principal Coordinate Analysis (PCoA) based on Bray–Curtis dissimilarities was used to visualize differences in viral and prokaryotic community composition, as well as functional gene profiles. The statistical significance of community clustering was assessed using the adonis2 function from the vegan package with 999 permutations 98 . Data availability The raw data for viral and prokaryotic genome, and source data used to generate figures for this study are available from https://figshare.com/s/d6199f1925be7033719e . Declarations Competing interests The authors declare that they have no competing interests. Author contributions Y.G.X., J.X.S., Y.G.W., K. A., and Z.S. H. conceived the study. Y.G.W., L.B.Y., and X. H.C performed the sample collection. J.X.S., Z.C., and S.L. performed the measurement of physiochemical parameters, DNA extraction. Y.G.X., J.X.S., Y.L.Q., Z.H.L., and Y.N.Q. performed the metagenomic analyses. Y.G.X., J.X.S., H.B.Z., Z.Z.Z., M.L. Y.G.W., K. A., and Z.S. H. wrote the manuscript. All authors discussed the results and commented on the manuscript. Acknowledgements This work was supported by grants from the National Natural Science Foundation of China (332471574, Z.S.H.; 32400002, Y.N.Q.; 42207145, Y.L.Q.; 42073079, Y.G.W.), and the National Key Basic Research Program of China (2015CB755903), and by the National Science Foundation under grant no. DBI2047598 (to K.A.). References Fuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399:541–548 Suttle CA (2007) Marine viruses-major players in the global ecosystem. Nat Rev Microbiol 5:801–812 Shiah FK, Lai CC, Chen TY, Ko CY, Tai JH, Chang CW (2022) Viral shunt in tropical oligotrophic ocean. Sci Adv 8:eabo2829 Suttle CA (2005) Viruses in the sea. Nature 437:356–361 Hurwitz BL, U'Ren JM (2016) Viral metabolic reprogramming in marine ecosystems. Curr Opin Microbiol 31:161–168 Ignacio-Espinoza JC, Ahlgren NA, Fuhrman JA (2020) Long-term stability and Red Queen-like strain dynamics in marine viruses. Nat Microbiol 5:265–271 Chevallereau A, Pons BJ, van Houte S, Westra ER (2022) Interactions between bacterial and phage communities in natural environments. Nat Rev Microbiol 20:49–62 Huang D et al (2024) Adaptive strategies and ecological roles of phages in habitats under physicochemical stress. Trends Microbiol 32:902–916 Thompson LR et al (2011) Phage auxiliary metabolic genes and the redirection of cyanobacterial host carbon metabolism. Proc. Natl. Acad. Sci. USA 108, E757–764 Kieft K et al (2021) Ecology of inorganic sulfur auxiliary metabolism in widespread bacteriophages. Nat Commun 12:3503 Lopez-Simon J, Vila-Nistal M, Rosenova A, De Corte D, Baltar F, Martinez-Garcia M (2023) Viruses under the Antarctic Ice Shelf are active and potentially involved in global nutrient cycles. Nat Commun 14:8295 Gazitua MC et al (2021) Potential virus-mediated nitrogen cycling in oxygen-depleted oceanic waters. ISME J 15:981–998 Stern RJ (2021) Ocean trenches. Encycl Geol 2:845–854 Jamieson AJ, Fujii T, Mayor DJ, Solan M, Priede IG (2010) Hadal trenches: the ecology of the deepest places on Earth. Trends Ecol Evol 25:190–197 Du M et al (2021) Geology, environment, and life in the deepest part of the world's oceans. Innovation 2:100109 Nunoura T et al (2015) Hadal biosphere: Insight into the microbial ecosystem in the deepest ocean on Earth. Proc. Natl. Acad. Sci. USA 112, E1230–E1236 Weston JNJ, Jensen EL, Hasoon MSR, Kitson JJN, Stewart HA, Jamieson AJ (2022) Barriers to gene flow in the deepest ocean ecosystems: Evidence from global population genomics of a cosmopolitan amphipod. Sci Adv 8:eabo6672 Hiraoka S et al (2020) Microbial community and geochemical analyses of trans-trench sediments for understanding the roles of hadal environments. ISME J 14:740–756 Dang YR et al (2024) Phytoplankton-derived polysaccharides and microbial peptidoglycans are key nutrients for deep-sea microbes in the Mariana Trench. Microbiome 12, 77 Zhou YL, Mara P, Cui GJ, Edgcomb VP, Wang Y (2022) Microbiomes in the Challenger Deep slope and bottom-axis sediments. Nat Commun 13:1515 Jian HH et al (2021) Diversity and distribution of viruses inhabiting the deepest ocean on Earth. ISME J 15:3094–3110 Zhou YL, Mara P, Vik D, Edgcomb VP, Sullivan MB, Wang Y (2022) Ecogenomics reveals viral communities across the Challenger Deep oceanic trench. Commun Biol 5:1055 Zhao J et al (2022) Novel viral communities potentially assisting in carbon, nitrogen, and sulfur metabolism in the upper slope sediments of Mariana Trench. mSystems 7, e0135821 Xiao X et al (2025) Microbial ecosystems and ecological driving forces in the deepest ocean sediments. Cell 188:1363–1377 Rogers AD (2015) Environmental change in the deep ocean. Annu Rev Env Resour 40:1–38 Ge W et al (2012) Oxygenase-catalyzed ribosome hydroxylation occurs in prokaryotes and humans. Nat Chem Biol 8:960–962 Van Staalduinen LM, Jia ZC (2015) Post-translational hydroxylation by 2OG/Fe(ll)-dependent oxygenases as a novel regulatory mechanism in bacteria. Front Microbi 5:798 Luo EL, Eppley JM, Romano AE, Mende DR, DeLong EF (2020) Double-stranded DNA virioplankton dynamics and reproductive strategies in the oligotrophic open ocean water column. ISME J 14:1304–1315 Suttle CA, Feng C (1992) Mechanisms and rates of decay of marine viruses in seawater. Appl Environ Microbil 58:3721–3729 Wommack KE, Hill RT, Muller TA, Colwell RR (1996) Effects of sunlight on bacteriophage viability and structure. Appl Environ Microbiol 62:1336–1341 Wilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea: viruses play critical roles in the structure and function of aquatic food webs. Bioscience 49:781–788 Middelboe M, Jorgensen NOG, Kroer N (1996) Effects of viruses on nutrient turnover and growth efficiency of noninfected marine bacterioplankton. Appl Environ Microbiol 62:1991–1997 Guidi L et al (2016) Plankton networks driving carbon export in the oligotrophic ocean. Nature 532:465–470 Laber CP et al (2018) facilitation of carbon export in the North Atlantic. Nat Microbiol 3:537–547 Sheyn U et al (2018) Expression profiling of host and virus during a coccolithophore bloom provides insights into the role of viral infection in promoting carbon export. ISME J 12:704–713 Zhang CL et al (2018) Evolving paradigms in biological carbon cycling in the ocean. Natl Sci Rev 5:481–499 Dunne JP, Sarmiento JL, Gnanadesikan A (2007) A synthesis of global particle export from the surface ocean and cycling through the ocean interior and on the seafloor. Global Biogeochem Cycles 21:GB4006 Arístegui J, Gasol JM, Duarte CM, Herndl GJ (2009) Microbial oceanography of the dark ocean's pelagic realm. Limnol Oceanogr 54:1501–1529 Dell'Anno A, Corinaldesi C, Danovaro R (2015) Virus decomposition provides an important contribution to benthic deep-sea ecosystem functioning. Proc. Natl. Acad. Sci. USA 112, E2014–E2019 Danovaro R et al (2008) Major viral impact on the functioning of benthic deep-sea ecosystems. Nature 454:1084–1087 Li Y et al (2014) Lytic viral infection of bacterioplankton in deep waters of the western Pacific Ocean. Biogeosciences 11:2531–2542 Howard-Varona C, Hargreaves KR, Abedon ST, Sullivan MB (2017) Lysogeny in nature: mechanisms, impact and ecology of temperate phages. ISME J 11:1511–1520 Turner D et al (2023) Abolishment of morphology-based taxa and change to binomial species names: 2022 taxonomy update of the ICTV bacterial viruses subcommittee. Arch Virol 168:74 Fuchsman CA et al (2019) Cyanobacteria and cyanophage contributions to carbon and nitrogen cycling in an oligotrophic oxygen-deficient zone. ISME J 13:2714–2726 Dieppa-Colón E, Martin C, Kosmopoulos JC, Anantharaman K (2025) Prophage-DB: a comprehensive database to explore diversity, distribution, and ecology of prophages. Environ Microbiome 20:1 Song XW et al (2025) Rhizosphere-triggered viral lysogeny mediates microbial metabolic reprogramming to enhance arsenic oxidation. Nat Commun 16:4048 Francis JD, Yanaç K, Uyaguari-Diaz MI (2025) Seasonal patterns of viromes in urban aquatic environments of Manitoba. Appl Environ Microbil 91:e00408–e00424 Du S et al (2025) Marine phages exhibit high genetic diversity and global-scale biogeographic patterns. Commun Biol 8:1240 Auladell A et al (2023) Seasonality of biogeochemically relevant microbial genes in a coastal ocean microbiome. Environ Microbiol 25:1465–1483 Haro-Moreno JM, Roda-Garcia JJ, Molina-Pardines C, López-Pérez M (2025) The hidden genetic reservoir: structural variants as drivers of marine microbial and viral microdiversity. Environ Microbiome 20:110 Huang YH, Zhang XX, Xin Y, Tian JW, Li M (2024) Distinct microbial nitrogen cycling processes in the deepest part of the ocean. Msystems 9:e00243–e00224 Gerrish PJ, Colato A, Sniegowski PD (2013) Genomic mutation rates that neutralize adaptive evolution and natural selection. J R Soc 10:20130329 Gerrish PJ, Colato A, Perelson AS, Sniegowski PD (2007) Complete genetic linkage can subvert natural selection. Proc. Natl. Acad. Sci. USA. 104, 6266–6271 Lynch M et al (2016) Genetic drift, selection and the evolution of the mutation rate. Nat Rev Genet 17:704–714 Zhong ZP et al (2023) Lower viral evolutionary pressure under stable versus fluctuating conditions in subzero Arctic brines. Microbiome 11:174 Danovaro R, Corinaldesi C, Dell'Anno A, Snelgrove PVR (2017) The deep-sea under global change. Curr Biol 27:R461–R465 Cui D, Sun CJ, Guo CA, Cao W, Jiang FH, Ding HB (2023) Characteristics of dissolved sugars in the Southern Yap Trench from sea surface to hadal zone. J Oceanol Limnol 41:2117–2133 Garcia-Garcia N, Tamames J, Linz AM, Pedros-Alio C, Puente-Sanchez F (2019) Microdiversity ensures the maintenance of functional microbial communities under changing environmental conditions. ISME J 13:2969–2983 Xie YG et al (2025) Lake salinization on the Qinghai-Tibetan Plateau alters viral community composition and lifestyles. Commun Earth Environ 6:49 Yi Y et al (2023) A systematic analysis of marine lysogens and proviruses. Nat Commun 14:6013 Li GL, Cortez MH, Dushoff J, Weitz JS (2020) When to be temperate: on the fitness benefits of lysis vs. lysogeny. Virus Evol 6:veaa042 Bobay LM, Rocha EPC, Touchon M (2013) The adaptation of temperate bacteriophages to their host genomes. Mol Biol Evol 30:737–751 Hackl T et al (2023) Novel integrative elements and genomic plasticity in ocean ecosystems. Cell 186:47–62 Touchon M, de Sousa JAM, Rocha EPC (2017) Embracing the enemy: the diversification of microbial gene repertoires by phage-mediated horizontal gene transfer. Curr Opin Microbiol 38:66–73 Wendling CC, Refardt D, Hall AR (2021) Fitness benefits to bacteria of carrying prophages and prophage-encoded antibiotic-resistance genes peak in different environments. Evolution 75:515–528 Labrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317–327 Wang W, Zhao H, Han GZ (2020) Host-virus arms races drive elevated adaptive evolution in viral receptors. J Virol 94:e00684–e00620 Gonzalez-Serrano R, Rosselli R, Roda-Garcia JJ, Martin-Cuadrado AB, Rodriguez-Valera F, Dunne M (2023) Distantly related bacteriophages share tail fibers exhibiting properties of transient chaperone caps. Nat Commun 14:6517 Cram JA et al (2015) Seasonal and interannual variability of the marine bacterioplankton community throughout the water column over ten years. ISME J 9:563–580 Zhang XX, Xu W, Liu Y, Cai MW, Luo ZH, Li M (2018) Metagenomics reveals microbial diversity and metabolic potentials of seawater and surface sediment from a hadal biosphere at the Yap Trench. Front Microbiol 9:2402 Chen SF, Zhou YQ, Chen YR, Gu J (2018) fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34:884–890 Bankevich A et al (2012) SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol 19:455–477 Kang DD et al (2019) MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ 7:e7359 Parks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW (2015) CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res 25:1043–1055 Chaumeil PA, Mussig AJ, Hugenholtz P, Parks DH (2022) GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics 38:5315–5316 Parks DH, Chuvochina M, Chaumeil PA, Rinke C, Mussig AJ, Hugenholtz P (2020) A complete domain-to-species taxonomy for Bacteria and Archaea. Nat Biotechnol 38:1079–1086 Qi YL et al (2024) Analysis of nearly 3000 archaeal genomes from terrestrial geothermal springs sheds light on interconnected biogeochemical processes. Nat Commun 15:4066 Kieft K, Zhou ZC, Anantharaman K (2020) VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences. Microbiome 8:90 Guo JR et al (2021) VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome 9:37 Roux S et al (2019) Minimum information about an uncultivated virus genome (MIUViG). Nat Biotechnol 37:29–37 Nayfach S, Camargo AP, Schulz F, Eloe-Fadrosh E, Roux S, Kyrpides NC (2021) CheckV assesses the quality and completeness of metagenome-assembled viral genomes. Nat Biotechnol 39:578–585 Fu LM, Niu BF, Zhu ZW, Wu ST, Li WZ (2012) CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics 28:3150–3152 Camargo AP et al (2023) IMG/VR v4: an expanded database of uncultivated virus genomes within a framework of extensive functional, taxonomic, and ecological metadata. Nucleic Acids Res 51:D733–D743 Gregory AC et al (2019) Marine DNA viral macro- and microdiversity from pole to pole. Cell 177:1109–1123 Wu SF et al (2021) DeePhage: distinguishing virulent and temperate phage-derived sequences in metavirome data with a deep learning approach. Gigascience 10:giab056 Shang JY, Jiang JZ, Sun YN (2021) Bacteriophage classification for assembled contigs using graph convolutional network. Bioinformatics 37:I25–I33 Hyatt D, Chen GL, Locascio PF, Land ML, Larimer FW, Hauser LJ (2010) Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 11:119 Buchfink B, Xie C, Huson DH (2015) Fast and sensitive protein alignment using DIANOND. Nat Methods 12:59–60 Huerta-Cepas J et al (2019) eggNOG 5.0: a hierarchical, functionally and phylogenetically annotated orthology resource based on 5090 organisms and 2502 viruses. Nucleic Acids Res 47:D309–D314 Aramaki T et al (2020) KofamKOALA: KEGG Ortholog assignment based on profile HMM and adaptive score threshold. Bioinformatics 36:2251–2252 Shaffer M et al (2020) DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res 48:8883–8900 Mirdita M, Schütze K, Moriwaki Y, Heo L, Ovchinnikov S, Steinegger M (2022) ColabFold: making protein folding accessible to all. Nat Methods 19:679–682 van Kempen M et al (2024) Fast and accurate protein structure search with Foldseek. Nat Biotechnol 42:243–246 Roux S et al (2023) iPHoP: An integrated machine learning framework to maximize host prediction for metagenome-derived viruses of archaea and bacteria. PLoS Biol 21:e3002083 Aroney STN, Newell RJP, Nissen JN, Camargo AP, Tyson GW, Woodcroft B (2025) CoverM: read alignment statistics for metagenomics. Bioinformatics 41:btaf147 Langdon WB (2015) Performance of genetic programming optimised Bowtie2 on genome comparison and analytic testing (GCAT) benchmarks. BioData Min 8:1 Olm MR, Crits-Christoph A, Bouma-Gregson K, Firek BA, Morowitz MJ, Banfield JF (2021) inStrain profiles population microdiversity from metagenomic data and sensitively detects shared microbial strains. Nat Biotechnol 39:727–736 Dixon P (2003) VEGAN, a package of R functions for community ecology. J Veg Sci 14:927–930 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.docx Supplementary Information Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8682536","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":584572496,"identity":"5ca42c71-e749-4e81-bc55-4facfed290aa","order_by":0,"name":"Karthik Anantharaman","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYBACCSBmZjBIkDNA8GEkbi2MzUAtxkAtjA0kaGFISNwA0YJsFw4g2X72+eOCgrT07eyHnz9gbLOw529gPnibB48WaZ50w+YZBjm5O3vSDBsY2yQSZxxgS7bGp0WOIY2xmcegInfDgRzGBsZtEgkGDDxm0ni18D8Da0k3OP8GrMXegIH/G14t0hJgW3ISDG5AbGHcwMDDhleL5IxnjLN5DNIMd854Zjgj8R/QL4fZjC3n4NEicT6N4TPPn2R5c/7kBx8+nKmz529vfnjjDR4tqCABRDATrXwUjIJRMApGAS4AAOosRIi/xeGTAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-9584-2491","institution":"University of Wisconsin-Madison","correspondingAuthor":true,"prefix":"","firstName":"Karthik","middleName":"","lastName":"Anantharaman","suffix":""},{"id":584572497,"identity":"21c0aa64-c8a4-4892-9b15-945b39bd68f7","order_by":1,"name":"Yuan-Guo Xie","email":"","orcid":"https://orcid.org/0009-0003-7760-3395","institution":"University of Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Yuan-Guo","middleName":"","lastName":"Xie","suffix":""},{"id":584572498,"identity":"051021c9-797f-4dc7-9df6-f37615b8e686","order_by":2,"name":"Jianxing Sun","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Jianxing","middleName":"","lastName":"Sun","suffix":""},{"id":584572499,"identity":"57345cf1-c913-4993-bac8-f7586a1e9809","order_by":3,"name":"Yanling Qi","email":"","orcid":"https://orcid.org/0009-0008-7543-6286","institution":"Shenzhen University","correspondingAuthor":false,"prefix":"","firstName":"Yanling","middleName":"","lastName":"Qi","suffix":""},{"id":584572500,"identity":"e32441db-8c8d-4dde-af19-5e102418de25","order_by":4,"name":"Libo Yu","email":"","orcid":"","institution":"Third Institute of Oceanography","correspondingAuthor":false,"prefix":"","firstName":"Libo","middleName":"","lastName":"Yu","suffix":""},{"id":584572501,"identity":"cdd1346a-12fa-486d-b900-17307004acc5","order_by":5,"name":"Hongbo Zhou","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Hongbo","middleName":"","lastName":"Zhou","suffix":""},{"id":584572502,"identity":"88dfd096-6638-4a53-9f8d-47d0469597fb","order_by":6,"name":"Zhu Chen","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Zhu","middleName":"","lastName":"Chen","suffix":""},{"id":584572503,"identity":"57fbceb5-b4e9-4e2a-95c1-2adbeeb43a05","order_by":7,"name":"Zhen-Hao Luo","email":"","orcid":"","institution":"University of Vienna","correspondingAuthor":false,"prefix":"","firstName":"Zhen-Hao","middleName":"","lastName":"Luo","suffix":""},{"id":584572504,"identity":"34239f1c-cdf4-457e-986d-c721ed5d8db5","order_by":8,"name":"Yan-Ni Qu","email":"","orcid":"","institution":"University of Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Yan-Ni","middleName":"","lastName":"Qu","suffix":""},{"id":584572505,"identity":"84289d55-a488-4699-af82-6f86c3376849","order_by":9,"name":"Li Shen","email":"","orcid":"https://orcid.org/0000-0001-9204-1114","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Shen","suffix":""},{"id":584572506,"identity":"e0976f4b-f1a7-4a04-a14f-bb3b3479038f","order_by":10,"name":"Xinhua Chen","email":"","orcid":"https://orcid.org/0000-0003-2379-7596","institution":"Fujian Agriculture and Forestry University","correspondingAuthor":false,"prefix":"","firstName":"Xinhua","middleName":"","lastName":"Chen","suffix":""},{"id":584572507,"identity":"07985ed5-d429-4965-b3de-9d1452a310e5","order_by":11,"name":"Zhichao Zhou","email":"","orcid":"https://orcid.org/0000-0002-9120-6762","institution":"Shenzhen University","correspondingAuthor":false,"prefix":"","firstName":"Zhichao","middleName":"","lastName":"Zhou","suffix":""},{"id":584572508,"identity":"88a8a0ef-3dad-40ef-851a-c0405baae388","order_by":12,"name":"Meng Li","email":"","orcid":"https://orcid.org/0000-0001-8675-0758","institution":"Shenzhen University","correspondingAuthor":false,"prefix":"","firstName":"Meng","middleName":"","lastName":"Li","suffix":""},{"id":584572509,"identity":"1c261986-2f76-4fbf-8413-c8d2dd167b67","order_by":13,"name":"Yuguang Wang","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Yuguang","middleName":"","lastName":"Wang","suffix":""},{"id":584572510,"identity":"bfbb92f4-b065-46c7-b42b-c1e59275f4e1","order_by":14,"name":"Zheng-Shuang Hua","email":"","orcid":"https://orcid.org/0000-0002-2405-4228","institution":"University of Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"Zheng-Shuang","middleName":"","lastName":"Hua","suffix":""}],"badges":[],"createdAt":"2026-01-23 21:30:47","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8682536/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8682536/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102535626,"identity":"932b0d74-2f78-4abf-a93d-29585ecf9e71","added_by":"auto","created_at":"2026-02-12 17:20:23","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":427327,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDepth-resolved overview of viral community structure and lifestyle dynamics across the Yap Trench water column.\u003c/strong\u003e (a) Taxonomic composition and relative abundance of viral communities across water column depths. (b) Principal coordinate analysis (PCoA) of viral community composition based on Bray–Curtis dissimilarities. (c) Canonical correspondence analysis (CCA) linking viral community structure to environmental parameters. (d) Depth-stratified patterns of viral diversity and lifestyle distribution. (e) Proportions of viruses assigned to different lifestyle strategies (virulent vs. temperate). (f) Abundance dynamics of virulent and temperate viruses along the vertical water column. Abundance was calculated as normalized coverage (Cov/bp), where Cov/bp denotes the average per-base read coverage.\u003c/p\u003e","description":"","filename":"11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/a5d37e9e1eb3084277455564.jpg"},{"id":102535628,"identity":"a8c44a90-2bd5-4385-ade4-45a50f72ee25","added_by":"auto","created_at":"2026-02-12 17:20:23","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":715902,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDepth-resolved dynamics of virus–host interactions.\u003c/strong\u003e (a) Vertical distribution of viral community composition based on abundance across sampling depths (b) Relative abundance of viruses with predicted hosts across the water column. Colors represent the taxonomic classification of host lineages at the phylum or class level. (c) Vertical distribution of microbial community composition based on abundance across sampling depths. (d) Multivariate associations among viral lineages, predicted host taxa, infection strategies, and sampling depth. Abundance was calculated as normalized coverage (Cov/bp), where Cov/bp denotes the average per-base read coverage.\u003c/p\u003e","description":"","filename":"12.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/de18c82ea7e5af30daa84799.jpg"},{"id":102746577,"identity":"624ed63e-d665-4a29-bb6c-b83a5c6f51b7","added_by":"auto","created_at":"2026-02-16 08:58:19","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1099171,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional roles and distribution patterns of virus-encoded AMGs.\u003c/strong\u003e (a) Conceptual diagram illustrating virus–host metabolic interactions mediated by AMGs. (b) Depth-stratified abundance profiles of AMGs identified in viral genomes across the Yap Trench water column. Gene abundance data were normalized across taxa, and \u003cem\u003ez\u003c/em\u003e scores were generated.\u003c/p\u003e","description":"","filename":"13.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/74cd4a13e3d4907f12b6f6ab.jpg"},{"id":102747261,"identity":"f1eebf31-7f0a-4fad-a777-bed7d770eef6","added_by":"auto","created_at":"2026-02-16 09:04:19","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":301859,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDepth-associated microevolutionary dynamics in water column viral communities.\u003c/strong\u003e (a) Depth-dependent distributions of viral mutation density. Density plots showing the distribution of single-nucleotide variants (SNVs) per kilobase across viral genomes from surface (S), mesopelagic (M), and deep (D) layers. Vertical dashed lines indicate median values for each layer. (b–d) Proportions of viral genomes exhibiting SNVs across depth layers, shown separately for all viruses (b), virulent viruses (c), and temperate viruses (d). (d) Depth-dependent distributions of viral pN/pS ratios. Density plots showing the distribution of nonsynonymous to synonymous polymorphism ratios (pN/pS) across viral genomes from surface (S), mesopelagic (M), and deep (D) layers. Vertical dashed lines indicate median values for each layer. (f–h) Fractions of positively selected genes among mutated genes across depths, partitioned by all viruses (f), virulent viruses (g), and temperate viruses (h), reflecting depth-stratified signatures of adaptive evolution. Initial p-values were computed using the Mann-Whitney \u003cem\u003eU\u003c/em\u003e test and were further adjusted for multiple testing using the Benjamini–Hochberg correction in R.\u003c/p\u003e","description":"","filename":"14.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/aaba39486b71a14adcd85d8f.jpg"},{"id":102535630,"identity":"2afafae1-1b78-4bd5-a881-441c5ce1e22b","added_by":"auto","created_at":"2026-02-12 17:20:23","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":832338,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDepth-stratified distribution and functional signatures of positively selected viral genes. \u003c/strong\u003e(a) Relative abundance of viral genomes carrying positively selected genes across depths, grouped by predicted host lineage. (b) Functional categorization of positively selected genes by depth, based on annotation-derived metabolic classifications. (c) Depth-specific functional profiles of positively selected AMGs. (d) Genomic maps of representative viral genomes containing positively selected genes from surface, mesopelagic, and deep layers. Abbreviations: PTP, phage tail protein; PPP, phage portal protein; PTF, Phage tail fiber protein; hsdM, type I restriction enzyme M protein; gp46, Straboviridae exonuclease subunit 2; BTR, Bacteriophage translational regulator; gp22, Straboviridae capsid assembly scaffolding protein; gp23, Straboviridae major head protein; gp2, Salasmaviridae DNA polymerase; mltC, peptidoglycan lytic transglycosylase C; purC, phosphoribosylaminoimidazole-succinocarboxamide synthase; ENDOV, endonuclease V; Pep, Peptidase; PTFA, Phage tail fiber assembly protein; GPO, Phage capsid scaffolding protein serine peptidase; PST, Phage small terminase; gp25, Straboviridae baseplate wedge protein; K26895, Marseillevirus highly derived D5-like helicase-primase; dcm; DNA (cytosine-5)-methyltransferase 1; PCP, peptidoglycan catabolic process; gin, Enterobacteriaceae phage serine recombinase; RecU; recombination protein U; PH, Phage holin.\u003c/p\u003e","description":"","filename":"15.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/e1be0df8d9d5ab1c8e0eee5e.jpg"},{"id":102746519,"identity":"7286ee92-39f4-4196-9ed6-c8cf84a13532","added_by":"auto","created_at":"2026-02-16 08:58:01","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":616993,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDepth-dependent viral strategies and ecological roles in the Yap Trench. \u003c/strong\u003eThis schematic summarizes depth‐dependent changes in viral abundance, environmental conditions, and virus–host interactions across the water column (0–6000 m). Surface and mesopelagic waters are characterized by high viral abundance dominated by lytic cyanophages (e.g., \u003cem\u003eKyanoviridae\u003c/em\u003e), promoting viral shunt–mediated recycling of C, N, S, and P and reinforcing kill‐the‐winner dynamics. With increasing depth, viral abundance declines and temperate viruses (e.g., \u003cem\u003ePeduoviridae\u003c/em\u003e) become more prevalent, consistent with piggyback‐the‐winner strategies, genome integration, and horizontal gene transfer. Together, depth‐associated physicochemical gradients shape viral life histories, microbial interactions, and evolutionary processes in the hadal ocean. Abbreviations: VA, viral abundance; TEMP, temperature; DOC, dissolved organic carbon; O₂, oxygen; C, carbon; N, nitrogen; S, sulfur; P, phosphorus; HGT, horizontal gene transfer.\u003c/p\u003e","description":"","filename":"16.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/0194ac633e713d12162ddf17.jpg"},{"id":102751835,"identity":"eb1688a7-c1b0-4a7c-9a69-18fc8513d9a2","added_by":"auto","created_at":"2026-02-16 09:27:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5032504,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/b1fe36c5-3a50-4f7c-a7f8-e0ee212bbcb9.pdf"},{"id":102746572,"identity":"538cad69-40c1-4e92-a9cc-c6e101639e7f","added_by":"auto","created_at":"2026-02-16 08:58:17","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1858879,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8682536/v1/72897ae669308a67d762174c.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Depth-governed ecological and evolutionary partitioning of ocean trench viromes","fulltext":[{"header":"Main","content":"\u003cp\u003eViruses are the most abundant biological entities in the ocean and exert profound influence over microbial dynamics and marine biogeochemical cycles\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. By inducing host lysis, they release large pools of organic and inorganic nutrients, fueling the microbial loop and sustaining trophic connectivity across pelagic food webs\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Beyond their role in biomass turnover, viruses also shape microbial evolution via horizontal gene transfer, including the delivery of auxiliary metabolic genes (AMGs) that reprogram host metabolism and mediate key geochemical transformations\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. These dual roles, regulating microbial population dynamics and modulating ecosystem function, position viruses as key ecological agents in global biogeochemical cycling\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Yet despite this ecological importance, much of our understanding of marine viruses stems from studies conducted in surface or euphotic waters\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. How viral communities adapt and function in deeper oceanic layers, especially in the hadal zone, remains poorly resolved, largely due to sampling and technical challenges.\u003c/p\u003e \u003cp\u003eOceanic trench represents one of the most isolated and extreme habitats on Earth, formed at tectonic subduction zones and descending to depths beyond 6,000 mbsl\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. These environments are characterized by intense hydrostatic pressure (\u0026gt;\u0026thinsp;1,000 atm), near-freezing temperatures, complete darkness, and severe nutrient limitation\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Despite these constraints, trench ecosystems harbor diverse and metabolically specialized microbial life\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Recent metagenomic studies have uncovered rich microbial and viral diversity in trench sediments\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, including novel AMGs linked to deep biosphere functions\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. Yet, viral communities in the overlying water column, especially transitions from epipelagic to mesopelagic and further to hadal zones, remain largely unexplored. This gap hinders our understanding of how viral diversity, function, and evolution are structured by depth-dependent environmental pressures.\u003c/p\u003e \u003cp\u003eTo address these gaps, we conducted depth-resolved metagenomic profiling of viral communities across the full vertical profile of the Yap Trench (137.52\u0026ndash;138.70 \u0026deg;E, 8.02\u0026ndash;9.90\u0026deg;N), spanning from surface (2 mbsl) to hadal depths (\u0026gt;\u0026thinsp;6,000 mbsl). Here, using depth-resolved metagenomics of 17 water samples spanning the entire Yap Trench water column, we investigated how environmental gradient\u0026rsquo;s structure viral community composition, mediate host interactions, and further shape unveil evolutionary trajectories across ocean depth. Our findings together establish a depth-stratified framework for understanding viral persistence, adaptation, and biogeochemical influence across Earth\u0026rsquo;s deepest marine environments.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eViral community profiles and physicochemical characteristics of oceanic trench viromes\u003c/h2\u003e \u003cp\u003eFollowing sample collection from the Yap Trench, physicochemical analyses revealed pronounced vertical gradients associated with increasing depth (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb). Temperature declined sharply from surface levels to the hadal zone, dropping from near 29\u0026deg;C to approximately 1.5\u0026deg;C. Dissolved oxygen followed a non-linear distribution, initially decreasing through the mesopelagic zone before gradually increasing in deeper layers. Dissolved organic carbon (DOC) exhibited a mid-depth maximum between 200 and 300 mbsl, while both surface and hadal layers maintained relatively low DOC concentrations. In contrast, nitrate exhibited elevated concentrations in the deep zones. Additional physicochemical parameters are detailed in Supplementary Data 1. Such depth-stratified environmental gradients are likely key drivers of viral community dynamics, particularly across the mesopelagic-to-hadal transition zone.\u003c/p\u003e \u003cp\u003eTo investigate depth-associated changes in viral community, we conducted a comprehensive analysis of 17 metagenomic datasets spanning the Yap Trench water column. A total of 11,184 viral contigs with length\u0026thinsp;\u0026ge;\u0026thinsp;10 Kbp were recovered, forming the basis for subsequent profiling. Clustering these sequences at a 95% average nucleotide identity threshold yielded 8,520 viral operational taxonomic units (vOTUs), serving as a comprehensive representation of the viral community structure throughout the water column (Fig. S2). Yet, only 3.6% and 9.9% of these vOTUs matched entries in the Global Ocean Virome database v2 and IMG/VR v4 databases, respectively (Fig. S3), underscoring the exceptional genomic novelty of the Yap Trench virome. Among them, 188 and 499 vOTUs were categorized as high-quality and medium-quality, respectively, based on completeness\u0026thinsp;\u0026ge;\u0026thinsp;90 and \u0026ge;\u0026thinsp;50 (Supplementary Data 2).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSharp decline in viral diversity below 500 mbsl marks a pivotal ecological boundary\u003c/h3\u003e\n\u003cp\u003eViral diversity, in terms of both richness and abundance, exhibited consistent vertical trends throughout the water column, peaking before a sharp decline in the mesopelagic zone (100\u0026ndash;300 mbsl), highlighting the pronounced vertical stratification of viral communities along the depth gradient (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea and \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). This vertical structuring was further supported by the distribution of vOTUs, the majority of which showed strong sample specificity (Fig. S4). Only a limited fraction of vOTUs was shared across multiple samples, underscoring the pronounced spatial heterogeneity and vertical differentiation of viral assemblages in the deep-sea water column. Of the 8,520 vOTUs identified, 5,742 (67.4%) could be classified into 99 viral families (Supplementary Data 2). Surface and mesopelagic communities predominated by \u003cem\u003eKyanoviridae\u003c/em\u003e (average 53.64%) and \u003cem\u003eAutographiviridae\u003c/em\u003e (average 9.46%), while deep-water assemblages (500\u0026ndash;6000 mbsl) were enriched in unclassified taxa (average 48.97%), alongside \u003cem\u003eHendrixvirinae\u003c/em\u003e (average 3.71%) and \u003cem\u003ePeduoviridae\u003c/em\u003e (average 12.93%). The sharp decline in viral abundance between mesopelagic and deeper waters was largely attributable to the marked reduction of \u003cem\u003eKyanoviridae\u003c/em\u003e in deep layers.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePrincipal coordinate analysis (PCoA) revealed three distinct clusters of viral communities stratified by depth: 2\u0026ndash;50 mbsl for surface zones, 100\u0026ndash;300 mbsl for mesopelagic zones, and 500\u0026ndash;6000 mbsl for deep zones, further corroborating the pronounced vertical structuring of these communities (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec and d; \u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.55, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Even without the consideration of \u003cem\u003eKyanoviridae\u003c/em\u003e, the viral communities still clustered into the same three groups (Fig. S5), suggesting both abundant and rare viruses affect the viral community composition across different layers. Canonical correspondence analysis (CCA) identified depth and salinity as the principal environmental axis structuring viral communities. These two factors show an approximately orthogonal relationship, with depth governing viral assemblages in deep layers and salinity shaping those in surface layers. In contrast, DOC was more closely aligned with mesopelagic communities, suggesting distinct resource-driven dynamics at intermediate depths (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). These relationships between environmental factors and viral communities mirrored, to a considerable extent, those observed for microbial communities (Fig. S6). Together, these data identify the mesopelagic-to-bathypelagic transition as a critical ecological inflection point in viral community structure, likely driven by a combination of abiotic and biotic factors.\u003c/p\u003e\n\u003ch3\u003eHost specificity and viral lifestyle underlie depth-specific community transitions\u003c/h3\u003e\n\u003cp\u003eTo uncover the biological mechanisms underlying these community shifts, we next examined viral lifestyles and host associations, aiming to identify how host biogeography and infection strategies influence viral persistence across depths. We identified 697 (7.89%) vOTUs as temperate viruses and 3,525 (43.43%) vOTUs as virulent viruses (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee; Supplementary Data 2). The richness of both temperate and virulent viruses peaked in the mesopelagic zones but declined sharply beyond the mesopelagic layers, with virulent viruses consistently maintaining overwhelming diversity across all depths (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Notably, the abundance of temperate viruses remained comparatively stable throughout the water column, whereas virulent viruses showed pronounced variability (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef). Their populations were elevated in the surface and mesopelagic zones, but declined sharply at greater depths. These patterns indicate that vertical changes in viral diversity are primarily driven by fluctuations in virulent viral populations. Such depth-dependent shifts in the balance between virulent and temperate viruses imply a transition in viral\u0026ndash;host strategies: lytic infections dominate in the shallower layers, consistent with kill-the-winner dynamics, whereas the deeper layers show a modest but detectable rise in temperate lifestyles, signaling a shift toward a piggyback-the-winner mode as host turnover declines (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef).\u003c/p\u003e \u003cp\u003eBy applying a computational host-prediction framework (see Methods), we assigned 1,760 vOTUs to putative microbial hosts (Supplementary Data 2). Viruses linked to \u003cem\u003eGammaproteobacteria\u003c/em\u003e, \u003cem\u003eAlphaproteobacteria\u003c/em\u003e, \u003cem\u003eBacteroidota\u003c/em\u003e, and \u003cem\u003eChloroflexota\u003c/em\u003e predominated in surface waters, while archaeal viruses, particularly those infecting \u003cem\u003eNitrososphaeria\u003c/em\u003e and \u003cem\u003ePoseidoniia\u003c/em\u003e, were largely confined to mesopelagic depths, in line with host distributions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea\u0026ndash;c). An exception was observed for the Cyanobacteria-Cyanophage pair: \u003cem\u003eCyanobacteria\u003c/em\u003e were most abundant in surface waters, but cyanophages peaked in the mesopelagic zone, indicating spatial decoupling between host and virus. Viral contigs from deep waters exhibited a higher rate of host assignment compared with surface and mesopelagic layers. This is likely caused by the simplified community structure dominated by \u003cem\u003eGammaproteobacteria\u003c/em\u003e and \u003cem\u003eFirmicutes\u003c/em\u003e. Among these, viruses infecting \u003cem\u003eGammaproteobacteria\u003c/em\u003e were consistently abundant across all depths, presumably due to the broad ecological range of their hosts. For instance, some lineages such as \u003cem\u003eAlteromonadaceae\u003c/em\u003e are ubiquitous throughout the water column, whereas others exhibit marked depth specificity (Fig. S7). It is noteworthy that most Gammaproteobacteria-infecting viruses were affiliated with \u003cem\u003eKyanoviridae\u003c/em\u003e and \u003cem\u003ePeduoviridae\u003c/em\u003e, representing the most abundant virulent and temperate lineages, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Both lineages also infect additional microbial groups, albeit with contrasting host ranges. \u003cem\u003eKyanoviridae\u003c/em\u003e display a broad host spectrum, extending to major surface and mesopelagic taxa such as \u003cem\u003eBacteroidota\u003c/em\u003e and \u003cem\u003eCyanobacteria\u003c/em\u003e. These host taxa peaked in the mesopelagic zone but declined sharply below 300 mbsl (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). In contrast, \u003cem\u003ePeduoviridae\u003c/em\u003e exhibit a narrower range, primarily targeting \u003cem\u003eGammaproteobacteria\u003c/em\u003e and \u003cem\u003eFirmicutes\u003c/em\u003e, two lineages that are particularly enriched in deep-sea environments. Together, these findings suggest that host biogeography and lifestyle strategy are key factors driving viral community turnover with depth.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eViruses reprogram host metabolism via depth-stratified AMGs\u003c/h3\u003e\n\u003cp\u003eTo determine whether viruses influence host function in a depth-dependent manner, we examined the distribution and taxonomic context of AMGs across the water column. Virally encoded AMGs span diverse host pathways involved in key biogeochemical cycles (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea), highlighting their roles in shaping microbial function and ecosystem processes. Viral AMGs displayed clear vertical stratification in both function and host association, reflecting depth-specific metabolic adaptation. In the mesopelagic zone (100\u0026ndash;300 mbsl), viruses exhibited a high diversity of AMGs involved in key metabolic functions such as carbohydrate utilization (\u003cem\u003egalE\u003c/em\u003e, \u003cem\u003eppdK\u003c/em\u003e, \u003cem\u003eglgM\u003c/em\u003e) and nitrogen and sulfur cycling (\u003cem\u003enorQ\u003c/em\u003e, \u003cem\u003enosD\u003c/em\u003e, \u003cem\u003ecycC\u003c/em\u003e, \u003cem\u003ecycH\u003c/em\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, Supplementary Data 3). The observed functional repertoire may in part be attributable to the richness of viral taxa in mid-water environments. These AMGs were primarily encoded by viruses infecting \u003cem\u003eBacteroidota\u003c/em\u003e, \u003cem\u003eChloroflexota\u003c/em\u003e, and \u003cem\u003eCyanobacteria\u003c/em\u003e, taxa known to be metabolically active at mid-depths. Notably, photosynthesis-related AMGs (e.g., \u003cem\u003epsbA\u003c/em\u003e and \u003cem\u003epetE\u003c/em\u003e) also peaked in the mesopelagic and were broadly associated with diverse microbial taxa. Phylogenetic placement showed that \u003cem\u003epsbA\u003c/em\u003e genes clustered with cyanophage homologs (Fig. S8), suggesting lateral gene transfer and broader dissemination beyond canonical cyanophages. This implies that viruses may help maintain host phototrophic near the lower edge of the photic zone, potentially extending photosynthetic functionality under low-light conditions. Additionally, AMGs related to nucleotide metabolism and cofactor biosynthesis were most enriched in mesopelagic zones, potentially facilitating host biosynthetic processes and promoting viral proliferation under moderately favorable energetic conditions.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn contrast, deep-sea viral communities (\u0026gt;\u0026thinsp;500 mbsl) exhibited a marked depletion of nutrient cycling AMGs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), potentially resulting from the limited viral\u0026ndash;host interactions characteristic of deep-sea environments. However, genes related to environmental sensing and cellular adaptation, such as signal transduction, transport, and stress response, were detected in deep-sea viral communities (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). These included two-component regulators (\u003cem\u003ecsrA\u003c/em\u003e, \u003cem\u003eyesN\u003c/em\u003e, \u003cem\u003enarL\u003c/em\u003e) and motility-related genes (\u003cem\u003epilT\u003c/em\u003e, \u003cem\u003ecpaB\u003c/em\u003e), largely encoded by viruses infecting \u003cem\u003eGammaproteobacteria\u003c/em\u003e and \u003cem\u003eFirmicutes\u003c/em\u003e. Such functions may enable hosts to perceive physicochemical stressors and optimize the allocation of scarce resources, enhancing fitness in energy-limited deep-sea environments. These findings suggest that viruses infecting \u003cem\u003eGammaproteobacteria\u003c/em\u003e and \u003cem\u003eFirmicutes\u003c/em\u003e may influence host metabolism and motility, potentially enhancing host fitness under deep-sea conditions. Collectively, these data reveal that viral AMGs are reflect a layered functional architecture that aligns with host ecology and environmental context. Viral reprogramming of host metabolism is most pronounced in the mesopelagic zone, where energy and nutrient gradients intersect, while deeper waters favor adaptation-related functions. This stratified pattern suggests that viruses play depth-specific roles in modulating host ecological performance across the oceanic water column.\u003c/p\u003e\n\u003ch3\u003eDepth imposes selective pressures shaping viral microevolution\u003c/h3\u003e\n\u003cp\u003eGiven the strong functional stratification of viral genes across depth layers, we further investigated whether such adaptations are evolutionarily conserved and under selective pressure. Specifically, we examined how mutation rates and selection regimes vary across ocean depths to reveal the microevolutionary processes underpinning viral persistence in distinct environments. With increasing depth, the marine environment becomes progressively more extreme, characterized by persistent low temperatures, high hydrostatic pressure, and severe energy and nutrient limitation\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. These physicochemical constraints likely impose intensified selective pressures, thereby contributing to the marked decline in viral diversity observed in deeper waters. To evaluate this hypothesis, we further examined the microevolutionary trajectories of viruses across different depth layers. Our results showed that the proportion of vOTUs carrying mutations was lower in the deep layer (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb; Supplementary Data 4), but the diversity of mutated nucleotides per loci tended to be higher (Fig. S9). Mutation density (SNVs per Kbp) also declined markedly with depth, with median values of 19.4 at the surface, 15.8 in the mesopelagic, and 12.0 in the deep ocean (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). This suggests the existence of different evolutionary trajectories of viruses across depth layers. Surface and mesopelagic viruses exhibited a larger pool of mutations, with more genomic sites affected. In contrast, deep-sea viruses accumulated fewer mutations overall, but individual sites displayed greater nucleotide diversity. This pattern may increase the likelihood of beneficial variants being preserved under specific evolutionary circumstances. Consistent with this, viruses displayed progressively relaxed purifying selection with depth, as indicated by increasing median pN/pS values, from 0.106 at the surface to 0.141 in the mesopelagic and 0.148 in deep waters (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). Moreover, the proportion of genes under positive selection was significantly higher in deep-sea viruses compared with those from the upper layers (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef). These trends remained robust when viruses were further divided by lifestyle into virulent and temperate groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, d, g and h). Taxonomically, these genes were associated with dominant microbial host lineages at each depth (Fig. S11a). In absolute numbers, however, virulent viruses peaked in the mesopelagic zone, whereas temperate viruses increased steadily with depth (Fig. S11b).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eBy examining individual viruses, we found that higher SNV counts were mostly associated with dominant viruses. Those infecting \u003cem\u003eAlphaproteobacteria\u003c/em\u003e, \u003cem\u003eGammaproteobacteria\u003c/em\u003e, \u003cem\u003eChloroflexota\u003c/em\u003e, \u003cem\u003eCyanobacteria\u003c/em\u003e, and \u003cem\u003eBacteroidota\u003c/em\u003e predominated in the surface and mesopelagic zones, whereas viruses targeting \u003cem\u003eGammaproteobacteria\u003c/em\u003e and \u003cem\u003eFirmicutes\u003c/em\u003e prevailed in deeper waters (Fig. S10a). Positively selected genes were also more frequently observed in these abundant viruses (Fig S11a), suggesting that positive selection strengthens their competitive capacity and further promotes their fitness and proliferation. This trend was particularly pronounced in deeper layers, where the proportion of viruses under positive selection increased with relative abundance\u0026mdash;from ~\u0026thinsp;20% in surface waters to ~\u0026thinsp;50% in the mesopelagic and ~\u0026thinsp;70% in the deep ocean (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). Functional annotation of these loci revealed depth-specific divergence shaped by environmental selection, though only 29.1% could be assigned putative functions (Fig. S12; Supplementary Data 5). Across all depths, structural and host-interaction genes consistently dominated, suggesting strong selection on viral entry, recognition, and infection, consistent with an ongoing virus\u0026ndash;host arms-race dynamics (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Lifestyle comparisons further showed that temperate viruses were enriched in genes related to genetic information processing and metabolism, whereas virulent viruses predominantly exhibited selection on host-recognition and structural assembly (Fig. S13), pointing to contrasting adaptive strategies between lifestyles.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe next examined positively selected AMGs to resolve how metabolic potential contributes to viral adaptation. The 2OG\u0026ndash;Fe(II) oxygenase gene was the most consistently enriched AMG across depths, peaking in the mesopelagic zone and largely encoded by viruses infecting depth-specific dominant microbial lineages (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, Fig. S14). Its enrichment likely reflects selective pressures favoring hydroxylation-mediated pathways that support host metabolism under nutrient- and oxygen-limited conditions\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Beyond this, functional divergence of AMGs was apparent across depths: nitrogen- and sulfur-cycling genes were enriched in the mesopelagic zone, suggesting viral enhancement of host biogeochemical roles, whereas deep-sea AMGs were dominated by nucleotide metabolism, such as cytosine methylation (DNA (cytosine-5)-methyltransferase, \u003cem\u003edcm\u003c/em\u003e) encoded largely by Gammaproteobacteria-infecting viruses (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Noticeably, most positively selected AMGs were encoded by virulent viruses across all depths (Fig. S15), whereas temperate virus-associated AMGs were largely restricted to the deep ocean and primarily linked to cofactor biosynthesis and nucleotide metabolism. Collectively, these results suggest that positively selected AMGs follow depth-stratified, lineage-specific, and lifestyle-dependent patterns, highlighting their potential role in shaping host adaptation and ecosystem functioning across the oceanic water column.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eViruses exert strong control over marine microbial processes, yet their ecological and evolutionary dynamics across depth remain poorly resolved due to deep-sea sampling challenges. Previous studies have largely focused on surface and midwater communities, leaving the deep ocean, particularly hadal zones, underexplored. By establishing a full-depth continuum from the epipelagic to the hadal zone in Yap Trench and applying high-resolution metagenomics and advanced bioinformatics, our study overcomes these constraints and provides a more holistic view of how depth shapes viral diversity, virus\u0026ndash;host interactions, and adaptive evolution. Viral diversity peaked in the mesopelagic zone, where DOC had the most prominent effect on their distribution. This could be attributed to the favorable conditions for virus-host interactions, characterized by moderate nutrient availability, reduced UV exposure, and a permissive thermal regime\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. The dominance of virulent over temperate viruses suggests a kill-the-winner dynamic and an efficient viral shunt, with continuous lytic turnover injecting bioavailable DOC\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e that fuels heterotrophic respiration and elevates oxygen demand. Although direct causality cannot be established, the concordance between DOC maxima, lytic dominance and steep O\u003csub\u003e2\u003c/sub\u003e decline beneath 200\u0026ndash;300 m suggests that viral lysis contributes to shaping the upper margin of the OMZ. This DOC\u0026ndash;lysis coupling provides a plausible mechanism linking mesopelagic DOC enrichment and viral activity to enhanced carbon recycling and oxygen drawdown in the ocean interior (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Field observations reinforce this interpretation, showing positive correlations between modeled carbon flux and viruses infecting phytoplankton and picoplankton, and demonstrating virus-induced aggregation and export of phytoplankton biomass into mesopelagic depths\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. This midwater peak contrasts with deeper layers, where viral diversity declines sharply. The reduction likely stems from limited flux of labile organic matter, as sinking particles take weeks to months to reach hadal depths and arrive largely refractory, restricting microbial and viral activity\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Harsh physicochemical conditions\u0026mdash;high pressure, low temperatures, and energy scarcity\u0026mdash;further constrain replication\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. Under such conditions, infrequent host encounters make lytic infection costly, favoring temperate strategies\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Virulent virus abundance thus decreases with depth, though their richness remains relatively high, suggesting many lineages persist in dormant or low-activity states with potential for reactivation\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Field observations support this hypothesis, showing detectable but reduced lytic activity in bathypelagic waters and a gradual virus-to-prokaryote ratio decline\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. By contrast, temperate viruses decline more slowly and can become relatively enriched in deep communities\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, reflecting a depth-driven shift from lytic dominance to lysogenic persistence.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eA major contributor to this vertical contrast is the family \u003cem\u003eKyanoviridae\u003c/em\u003e. Dominant in surface and mesopelagic waters but nearly absent in the deep ocean, \u003cem\u003eKyanoviridae\u003c/em\u003e largely account for the difference in community composition between midwater and hadal layers. Although typically recognized as cyanophages infecting \u003cem\u003eProchlorococcus\u003c/em\u003e and \u003cem\u003eSynechococcus\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, their unexpectedly high abundance in the mesopelagic layer (200\u0026ndash;300 m), where phototrophic cyanobacteria are relatively scarce, can be explained by multiple processes. A legacy effect may contribute, with large numbers of viral particles produced in surface waters transported downward through water column mixing and particle flux. The mesopelagic zone also often coincides with the deep chlorophyll maximum or the upper boundary of the oxygen minimum zone, where residual cyanobacterial activity continues to support viral replication\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. In addition, some \u003cem\u003eKyanoviridae\u003c/em\u003e may possess broader host ranges, potentially infecting additional bacterial lineages beyond cyanobacteria (e.g. \u003cem\u003ePseudomonadota\u003c/em\u003e, and \u003cem\u003eActinomycetota\u003c/em\u003e)\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Enhanced microbial respiration and organic matter availability in this layer foster a metabolically active microbial community, which indirectly sustains \u003cem\u003eKyanoviridae\u003c/em\u003e persistence even where phototrophic cyanobacteria are scarce, thereby explaining their high abundance in the mesopelagic zone compared with the deep ocean. Below the mesopelagic, viruses infecting \u003cem\u003eGammaproteobacteria\u003c/em\u003e become increasingly dominant, reaching their highest prevalence in the deep ocean, in contrast to the cyanophage-dominated upper layers. This vertical structuring reflects both host turnover and shifts in infection strategies. In nutrient-rich surface and mesopelagic waters, \u003cem\u003eKyanoviridae\u003c/em\u003e and \u003cem\u003eAutographiviridae\u003c/em\u003e prevail, with lytic infections fueling rapid host turnover and organic matter recycling, consistent with the dynamic nature of upper-ocean ecosystems\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. These viruses are chiefly associated with the D2472 and TMED112 families, SAR86-lineage \u003cem\u003eGammaproteobacteria\u003c/em\u003e characterized by extreme genomic streamlining, proteorhodopsin-enabled photoheterotrophy, and enhanced DOM-scavenging capacities, well adapted to upper-ocean environments\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. At greater depths, however, temperate \u003cem\u003ePeduoviridae\u003c/em\u003e become prominent (Fig. S6a), echoing reports of lysogeny as a hallmark of energy-limited deep-sea communities \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Their hosts, mainly including \u003cem\u003eAlcanivoracaceae\u003c/em\u003e and \u003cem\u003eOleiphilaceae\u003c/em\u003e, are lineages specialized for survival under energy-limited deep-ocean conditions\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e, where lysogeny provides a selective advantage against resource scarcity. Thus, depth-related transitions in Gammaproteobacteria-virus interactions encapsulate the broader ecological restructuring that shapes the microbial life throughout the water column.\u003c/p\u003e \u003cp\u003eOur results support a unified \u0026ldquo;dual-filter\u0026rdquo; framework in which depth governs both ecological opportunities and evolutionary trajectories of marine viruses, linking environmental gradients to viral adaptation. Depth-stratified viruses follow distinct evolutionary trajectories. Notably, elevated mutation rates in surface and mesopelagic layers did not translate into higher pN/pS values, indicating limited adaptive substitutions. This paradox can be explained by mutation\u0026ndash;selection dynamics: while increased mutation rates raise the chance of beneficial variants; excessive mutational input generates a heavy deleterious load that erodes the fitness advantage of adaptive alleles\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. Stronger purifying selection in surface waters further supports this view, suggesting that most mutations were deleterious. As a result, natural selection becomes less efficient, and the proportion of positively selected substitutions remains low despite abundant mutational supply. Evidence points to ecological variability as a potential driver of viral microevolution\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e. Compared to surface and mesopelagic layers, deep layers are more extreme with reduced light, lower temperature, higher pressure and limited nutrients\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. These conditions favor a shift in viral lifestyles from virulent to temperate, allowing better adaptation to adverse environments\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e. Temperate viruses can accelerate adaptive change via long-term host association, frequent gene acquisition and recombination, and carriage of accessory genes that are exposed to selection, which has been observed experimentally and in large-scale genomic surveys\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. Strikingly, the most direct evidence for this process in the deep ocean is the pervasive positive selection on integrase genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Such selection likely enhances integration efficiency, thereby stabilizing lysogeny and boosting host fitness under conditions of low cell density and chronic energy limitation\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e. This evolutionary signal underscores the pivotal role of temperate strategies in driving horizontal gene transfer, positioning lysogeny as a major force in restructuring host genomes and promoting long-term adaptability in deep-sea ecosystems\u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e. Moreover, adsorption- and entry-related structural genes (for example, tail and tail-fiber loci) also experienced positive selection (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Even in lysogeny-dominated deep-sea communities, temperate phages must occasionally re-enter transmission phases where adsorption efficiency and host range directly determine fitness\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Tail/tail-fiber receptor-binding proteins, particularly at ligand-contact tips, thus exhibit episodic diversifying selection. Mechanistically, this is expected, since these modules operate at the frontline of host recognition and immune evasion\u003csup\u003e\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e. These dynamics suggest that even in energy-limited deep waters, sporadic but critical host encounters within microhabitats sustain antagonistic coevolution, sustaining adaptive evolution in entry modules alongside the dominance of lysogeny.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eSampling, physicochemical assay, DNA extraction and metagenomic sequencing\u003c/h2\u003e \u003cp\u003eSeawater samples were collected from seventeen depths ranging from 2 mbsl to 6,000 mbsl above the Yap Trench during research cruises in June 2017. At each depth, approximately 10 L of seawater was filtered through 0.22 \u0026micro;m pore-size membrane filters (Millipore, MA, USA) to capture microbial biomass. Filters were immediately flash-frozen in liquid nitrogen and stored at \u0026minus;\u0026thinsp;80\u0026deg;C until further processing. Environmental parameters, including temperature, salinity, dissolved oxygen, and nutrient concentrations, were measured following established protocols\u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e and are summarized in Supplementary Data 1. Total DNA was extracted from filters using the PowerSoil DNA Isolation Kit (MoBio Laboratories, Carlsbad, CA, USA), following the manufacturer\u0026rsquo;s instructions. Metagenomic libraries were prepared and sequenced on an Illumina MiSeq platform using 2 \u0026times; 150 bp paired-end reads, generating an average of ~\u0026thinsp;38 Gbp of raw data per sample. Two additional metagenomic datasets generated from the same project were integrated for further analyses\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eAssembly and binning of metagenomics\u003c/h2\u003e \u003cp\u003eMetagenomic raw reads were quality filtered using fastp (v0.23.2) with parameters \u0026ldquo;q 20 -u 20 -l 50\u0026rdquo; to eliminate adapter sequence and low-quality bases\u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e. Filtered reads were assembled using SPAdes (v3.15.2) with the parameters \u0026ldquo;-k 21, 33, 55, 77, 99,127 -meta\u0026rdquo;\u003csup\u003e72\u003c/sup\u003e. Scaffolds from each assembled sample were aligned against the cleaned reads from all depth layers using BBMap (v38.92; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://sourceforge.net/projects/bbmap/\u003c/span\u003e\u003cspan address=\"http://sourceforge.net/projects/bbmap/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) with parameters \u0026ldquo;k\u0026thinsp;=\u0026thinsp;15 minid\u0026thinsp;=\u0026thinsp;0.9 build\u0026thinsp;=\u0026thinsp;1\u0026rdquo;. Scaffold coverage across samples was calculated using the jgi_summarize_bam_contig_depths script in MetaBAT (v2.12.1). Genome binning for each sample was performed on scaffolds using MetaBAT (v2.12.1), based on differential coverage information across depth layers\u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e. The completeness and contamination of MAGs were assessed using CheckM (v1.1.3)\u003csup\u003e74\u003c/sup\u003e. The taxonomy of MAGs was obtained using GTDB-tk (v2.3.0) with reference to the genome taxonomy database (GTDB; release 214)\u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e. The universal prokaryotic marker gene Ribosomal Protein S3 (\u003cem\u003erpS3\u003c/em\u003e) was identified following established methods\u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of viral genomes\u003c/h2\u003e \u003cp\u003ePutative viral contigs were identified using VIBRANT (v1.2.1) and VirSorter2 (v2.2.3), both run with default parameters\u003csup\u003e\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e\u003c/sup\u003e. For VirSorter2, only contigs with max_score\u0026thinsp;\u0026ge;\u0026thinsp;0.90 were retained as high-confidence viral sequences. In accordance with the minimum information about an uncultivated virus genome (MIUViG) standard\u003csup\u003e\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e\u003c/sup\u003e, viral contigs\u0026thinsp;\u0026le;\u0026thinsp;10 kb in length were excluded from downstream analyses. CheckV (v0.8.1) was used to assess the completeness and contamination of viral genomes, and those lacking viral genes were subsequently removed\u003csup\u003e\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e\u003c/sup\u003e. vOTUs were defined by clustering contigs at 95% nucleotide identity across 85% alignment fraction using CD-HIT-EST (v4.8.1) with parameters -c 0.95 -aS 0.85\u003csup\u003e82\u003c/sup\u003e. To assess the similarity between Yap Trench viruses and previously characterized marine viromes, we retrieved reference datasets from the IMG/VR v4\u003csup\u003e83\u003c/sup\u003e and Global Ocean Viromes 2.0 (GOV2) databases\u003csup\u003e\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e\u003c/sup\u003e. Viral lifestyle predictions were performed using DeePhage (v1.0), with probability scores\u0026thinsp;\u0026le;\u0026thinsp;0.20 interpreted as temperate and \u0026ge;\u0026thinsp;0.80 as virulent lifestyles\u003csup\u003e\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e\u003c/sup\u003e. Taxonomic classification of viral genomes was performed using PhaGCN2.0\u003csup\u003e86\u003c/sup\u003e, a deep learning-based tool for virus taxonomic assignment. Open reading frames (ORFs) were predicted using Prodigal (v2.6.3) in metagenomic mode (-p meta)\u003csup\u003e\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e\u003c/sup\u003e. Predicted proteins were annotated against multiple databases, including NCBI-nr, KEGG and EggNOG v5.0 using DIAMOND (v2.0.14.152) with an E-value threshold of 1e\u0026thinsp;\u0026minus;\u0026thinsp;5\u003csup\u003e88, 89 90\u003c/sup\u003e. Furthermore, DRAM-v.py (v1.5.0) was employed for functional annotation to refine the understanding of viral-encoded metabolic functions\u003csup\u003e\u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e\u003c/sup\u003e. For viral genes under positive selection but lacking functional annotations, protein structures were predicted using ColabFold (v1.5.2)\u003csup\u003e92\u003c/sup\u003e, and subsequent structural comparisons were performed with Foldseek (10-941cd33) to infer potential biological roles\u003csup\u003e\u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e\u003c/sup\u003e. Viral host prediction was conducted using the iPHoP pipeline with default parameter\u003csup\u003e\u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e\u003c/sup\u003e. To improve prediction resolution, the reference database was extended to include 734 MAGs from this study in combination with the native iPHoP database.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCalculating microbial and viral abundance\u003c/h2\u003e \u003cp\u003eThe abundance of vOTUs and rpS3 genes was quantified using average number of aligned reads overlapping each position (trimmed_mean) metric. Clean reads were mapped to datasets using CoverM (v0.6.1) with \u0026ldquo;make\u0026rdquo; parameters to generate alignment BAM files\u003csup\u003e\u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e\u003c/sup\u003e. Reads not meeting alignment criteria were filtered using \u0026ldquo;filter\u0026rdquo; module with the following parameters: \u0026ldquo;--min-read-percent-identity 95 --min-read-aligned-percent 75\u0026rdquo;. Final abundance values were calculated using the contig module with trimming and alignment thresholds set as \u0026ldquo;--trim-min 0.10 --trim-max 0.90 --min-read-percent-identity 0.95 --min-read-aligned-percent 0.75\u0026rdquo;. The abundance of each vOTUs or gene was estimated as the average per-base read coverage, normalized by the total read count of the corresponding library, and scaled to the mean read count across all 17 libraries.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eAnalysis of viral populations\u003c/h2\u003e \u003cp\u003eClean reads from each sample were mapped to the concatenated set of all representative vOTUs using Bowtie2 with default parameters\u003csup\u003e\u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e\u003c/sup\u003e. Single nucleotide variants were identified using inStrain (v1.5.4) in --database-mode with default parameters\u003csup\u003e\u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e\u003c/sup\u003e. SNV calling required a minimum site coverage of 5\u0026times;, and only variants with a minor allele frequency\u0026thinsp;\u0026ge;\u0026thinsp;5% were retained. To reduce false positives, variant alleles were required to exceed the expected Illumina sequencing error rate (1 \u0026times; 10⁻⁶). Genome-wide nucleotide diversity (SNVs per kilobase) and the ratio of nonsynonymous to synonymous substitutions (pN/pS) were calculated at both the genome and gene levels. The pN/pS ratio was used as a proxy for selective pressure, where values\u0026thinsp;\u0026gt;\u0026thinsp;1 indicate potential positive selection, and values\u0026thinsp;\u0026lt;\u0026thinsp;1 reflect purifying selection.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eAll statistical analyses were conducted in R (v4.1.3). Principal Coordinate Analysis (PCoA) based on Bray\u0026ndash;Curtis dissimilarities was used to visualize differences in viral and prokaryotic community composition, as well as functional gene profiles. The statistical significance of community clustering was assessed using the adonis2 function from the vegan package with 999 permutations\u003csup\u003e\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eThe raw data for viral and prokaryotic genome, and source data used to generate figures for this study are available from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://figshare.com/s/d6199f1925be7033719e\u003c/span\u003e\u003cspan address=\"https://figshare.com/s/d6199f1925be7033719e\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eCompeting interests\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor contributions\u003c/h2\u003e \u003cp\u003eY.G.X., J.X.S., Y.G.W., K. A., and Z.S. H. conceived the study. Y.G.W., L.B.Y., and X. H.C performed the sample collection. J.X.S., Z.C., and S.L. performed the measurement of physiochemical parameters, DNA extraction. Y.G.X., J.X.S., Y.L.Q., Z.H.L., and Y.N.Q. performed the metagenomic analyses. Y.G.X., J.X.S., H.B.Z., Z.Z.Z., M.L. Y.G.W., K. A., and Z.S. H. wrote the manuscript. All authors discussed the results and commented on the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThis work was supported by grants from the National Natural Science Foundation of China (332471574, Z.S.H.; 32400002, Y.N.Q.; 42207145, Y.L.Q.; 42073079, Y.G.W.), and the National Key Basic Research Program of China (2015CB755903), and by the National Science Foundation under grant no. DBI2047598 (to K.A.).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eFuhrman JA (1999) Marine viruses and their biogeochemical and ecological effects. Nature 399:541\u0026ndash;548\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuttle CA (2007) Marine viruses-major players in the global ecosystem. Nat Rev Microbiol 5:801\u0026ndash;812\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShiah FK, Lai CC, Chen TY, Ko CY, Tai JH, Chang CW (2022) Viral shunt in tropical oligotrophic ocean. Sci Adv 8:eabo2829\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuttle CA (2005) Viruses in the sea. Nature 437:356\u0026ndash;361\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHurwitz BL, U'Ren JM (2016) Viral metabolic reprogramming in marine ecosystems. Curr Opin Microbiol 31:161\u0026ndash;168\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIgnacio-Espinoza JC, Ahlgren NA, Fuhrman JA (2020) Long-term stability and Red Queen-like strain dynamics in marine viruses. Nat Microbiol 5:265\u0026ndash;271\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChevallereau A, Pons BJ, van Houte S, Westra ER (2022) Interactions between bacterial and phage communities in natural environments. Nat Rev Microbiol 20:49\u0026ndash;62\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang D et al (2024) Adaptive strategies and ecological roles of phages in habitats under physicochemical stress. Trends Microbiol 32:902\u0026ndash;916\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThompson LR et al (2011) Phage auxiliary metabolic genes and the redirection of cyanobacterial host carbon metabolism. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e 108, E757\u0026ndash;764\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKieft K et al (2021) Ecology of inorganic sulfur auxiliary metabolism in widespread bacteriophages. Nat Commun 12:3503\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLopez-Simon J, Vila-Nistal M, Rosenova A, De Corte D, Baltar F, Martinez-Garcia M (2023) Viruses under the Antarctic Ice Shelf are active and potentially involved in global nutrient cycles. Nat Commun 14:8295\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGazitua MC et al (2021) Potential virus-mediated nitrogen cycling in oxygen-depleted oceanic waters. ISME J 15:981\u0026ndash;998\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStern RJ (2021) Ocean trenches. Encycl Geol 2:845\u0026ndash;854\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJamieson AJ, Fujii T, Mayor DJ, Solan M, Priede IG (2010) Hadal trenches: the ecology of the deepest places on Earth. Trends Ecol Evol 25:190\u0026ndash;197\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDu M et al (2021) Geology, environment, and life in the deepest part of the world's oceans. Innovation 2:100109\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNunoura T et al (2015) Hadal biosphere: Insight into the microbial ecosystem in the deepest ocean on Earth. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e 112, E1230\u0026ndash;E1236\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeston JNJ, Jensen EL, Hasoon MSR, Kitson JJN, Stewart HA, Jamieson AJ (2022) Barriers to gene flow in the deepest ocean ecosystems: Evidence from global population genomics of a cosmopolitan amphipod. Sci Adv 8:eabo6672\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHiraoka S et al (2020) Microbial community and geochemical analyses of trans-trench sediments for understanding the roles of hadal environments. ISME J 14:740\u0026ndash;756\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDang YR et al (2024) Phytoplankton-derived polysaccharides and microbial peptidoglycans are key nutrients for deep-sea microbes in the Mariana Trench. \u003cem\u003eMicrobiome\u003c/em\u003e 12, 77\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou YL, Mara P, Cui GJ, Edgcomb VP, Wang Y (2022) Microbiomes in the Challenger Deep slope and bottom-axis sediments. Nat Commun 13:1515\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJian HH et al (2021) Diversity and distribution of viruses inhabiting the deepest ocean on Earth. ISME J 15:3094\u0026ndash;3110\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou YL, Mara P, Vik D, Edgcomb VP, Sullivan MB, Wang Y (2022) Ecogenomics reveals viral communities across the Challenger Deep oceanic trench. Commun Biol 5:1055\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao J et al (2022) Novel viral communities potentially assisting in carbon, nitrogen, and sulfur metabolism in the upper slope sediments of Mariana Trench. \u003cem\u003emSystems\u003c/em\u003e 7, e0135821\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXiao X et al (2025) Microbial ecosystems and ecological driving forces in the deepest ocean sediments. Cell 188:1363\u0026ndash;1377\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogers AD (2015) Environmental change in the deep ocean. Annu Rev Env Resour 40:1\u0026ndash;38\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGe W et al (2012) Oxygenase-catalyzed ribosome hydroxylation occurs in prokaryotes and humans. Nat Chem Biol 8:960\u0026ndash;962\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Staalduinen LM, Jia ZC (2015) Post-translational hydroxylation by 2OG/Fe(ll)-dependent oxygenases as a novel regulatory mechanism in bacteria. Front Microbi 5:798\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo EL, Eppley JM, Romano AE, Mende DR, DeLong EF (2020) Double-stranded DNA virioplankton dynamics and reproductive strategies in the oligotrophic open ocean water column. ISME J 14:1304\u0026ndash;1315\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuttle CA, Feng C (1992) Mechanisms and rates of decay of marine viruses in seawater. Appl Environ Microbil 58:3721\u0026ndash;3729\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWommack KE, Hill RT, Muller TA, Colwell RR (1996) Effects of sunlight on bacteriophage viability and structure. Appl Environ Microbiol 62:1336\u0026ndash;1341\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilhelm SW, Suttle CA (1999) Viruses and nutrient cycles in the sea: viruses play critical roles in the structure and function of aquatic food webs. Bioscience 49:781\u0026ndash;788\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiddelboe M, Jorgensen NOG, Kroer N (1996) Effects of viruses on nutrient turnover and growth efficiency of noninfected marine bacterioplankton. Appl Environ Microbiol 62:1991\u0026ndash;1997\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuidi L et al (2016) Plankton networks driving carbon export in the oligotrophic ocean. Nature 532:465\u0026ndash;470\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaber CP et al (2018) facilitation of carbon export in the North Atlantic. Nat Microbiol 3:537\u0026ndash;547\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheyn U et al (2018) Expression profiling of host and virus during a coccolithophore bloom provides insights into the role of viral infection in promoting carbon export. ISME J 12:704\u0026ndash;713\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang CL et al (2018) Evolving paradigms in biological carbon cycling in the ocean. Natl Sci Rev 5:481\u0026ndash;499\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDunne JP, Sarmiento JL, Gnanadesikan A (2007) A synthesis of global particle export from the surface ocean and cycling through the ocean interior and on the seafloor. Global Biogeochem Cycles 21:GB4006\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAr\u0026iacute;stegui J, Gasol JM, Duarte CM, Herndl GJ (2009) Microbial oceanography of the dark ocean's pelagic realm. Limnol Oceanogr 54:1501\u0026ndash;1529\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDell'Anno A, Corinaldesi C, Danovaro R (2015) Virus decomposition provides an important contribution to benthic deep-sea ecosystem functioning. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e 112, E2014\u0026ndash;E2019\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDanovaro R et al (2008) Major viral impact on the functioning of benthic deep-sea ecosystems. Nature 454:1084\u0026ndash;1087\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Y et al (2014) Lytic viral infection of bacterioplankton in deep waters of the western Pacific Ocean. Biogeosciences 11:2531\u0026ndash;2542\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward-Varona C, Hargreaves KR, Abedon ST, Sullivan MB (2017) Lysogeny in nature: mechanisms, impact and ecology of temperate phages. ISME J 11:1511\u0026ndash;1520\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTurner D et al (2023) Abolishment of morphology-based taxa and change to binomial species names: 2022 taxonomy update of the ICTV bacterial viruses subcommittee. Arch Virol 168:74\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFuchsman CA et al (2019) Cyanobacteria and cyanophage contributions to carbon and nitrogen cycling in an oligotrophic oxygen-deficient zone. ISME J 13:2714\u0026ndash;2726\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDieppa-Col\u0026oacute;n E, Martin C, Kosmopoulos JC, Anantharaman K (2025) Prophage-DB: a comprehensive database to explore diversity, distribution, and ecology of prophages. Environ Microbiome 20:1\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong XW et al (2025) Rhizosphere-triggered viral lysogeny mediates microbial metabolic reprogramming to enhance arsenic oxidation. Nat Commun 16:4048\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrancis JD, Yana\u0026ccedil; K, Uyaguari-Diaz MI (2025) Seasonal patterns of viromes in urban aquatic environments of Manitoba. Appl Environ Microbil 91:e00408\u0026ndash;e00424\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDu S et al (2025) Marine phages exhibit high genetic diversity and global-scale biogeographic patterns. Commun Biol 8:1240\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAuladell A et al (2023) Seasonality of biogeochemically relevant microbial genes in a coastal ocean microbiome. Environ Microbiol 25:1465\u0026ndash;1483\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaro-Moreno JM, Roda-Garcia JJ, Molina-Pardines C, L\u0026oacute;pez-P\u0026eacute;rez M (2025) The hidden genetic reservoir: structural variants as drivers of marine microbial and viral microdiversity. Environ Microbiome 20:110\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang YH, Zhang XX, Xin Y, Tian JW, Li M (2024) Distinct microbial nitrogen cycling processes in the deepest part of the ocean. Msystems 9:e00243\u0026ndash;e00224\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerrish PJ, Colato A, Sniegowski PD (2013) Genomic mutation rates that neutralize adaptive evolution and natural selection. J R Soc 10:20130329\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerrish PJ, Colato A, Perelson AS, Sniegowski PD (2007) Complete genetic linkage can subvert natural selection. \u003cem\u003eProc. Natl. Acad. Sci. USA.\u003c/em\u003e 104, 6266\u0026ndash;6271\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLynch M et al (2016) Genetic drift, selection and the evolution of the mutation rate. Nat Rev Genet 17:704\u0026ndash;714\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhong ZP et al (2023) Lower viral evolutionary pressure under stable versus fluctuating conditions in subzero Arctic brines. Microbiome 11:174\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDanovaro R, Corinaldesi C, Dell'Anno A, Snelgrove PVR (2017) The deep-sea under global change. Curr Biol 27:R461\u0026ndash;R465\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui D, Sun CJ, Guo CA, Cao W, Jiang FH, Ding HB (2023) Characteristics of dissolved sugars in the Southern Yap Trench from sea surface to hadal zone. J Oceanol Limnol 41:2117\u0026ndash;2133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarcia-Garcia N, Tamames J, Linz AM, Pedros-Alio C, Puente-Sanchez F (2019) Microdiversity ensures the maintenance of functional microbial communities under changing environmental conditions. ISME J 13:2969\u0026ndash;2983\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie YG et al (2025) Lake salinization on the Qinghai-Tibetan Plateau alters viral community composition and lifestyles. Commun Earth Environ 6:49\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYi Y et al (2023) A systematic analysis of marine lysogens and proviruses. Nat Commun 14:6013\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi GL, Cortez MH, Dushoff J, Weitz JS (2020) When to be temperate: on the fitness benefits of lysis vs. lysogeny. Virus Evol 6:veaa042\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBobay LM, Rocha EPC, Touchon M (2013) The adaptation of temperate bacteriophages to their host genomes. Mol Biol Evol 30:737\u0026ndash;751\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHackl T et al (2023) Novel integrative elements and genomic plasticity in ocean ecosystems. Cell 186:47\u0026ndash;62\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTouchon M, de Sousa JAM, Rocha EPC (2017) Embracing the enemy: the diversification of microbial gene repertoires by phage-mediated horizontal gene transfer. Curr Opin Microbiol 38:66\u0026ndash;73\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWendling CC, Refardt D, Hall AR (2021) Fitness benefits to bacteria of carrying prophages and prophage-encoded antibiotic-resistance genes peak in different environments. Evolution 75:515\u0026ndash;528\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLabrie SJ, Samson JE, Moineau S (2010) Bacteriophage resistance mechanisms. Nat Rev Microbiol 8:317\u0026ndash;327\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang W, Zhao H, Han GZ (2020) Host-virus arms races drive elevated adaptive evolution in viral receptors. J Virol 94:e00684\u0026ndash;e00620\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonzalez-Serrano R, Rosselli R, Roda-Garcia JJ, Martin-Cuadrado AB, Rodriguez-Valera F, Dunne M (2023) Distantly related bacteriophages share tail fibers exhibiting properties of transient chaperone caps. Nat Commun 14:6517\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCram JA et al (2015) Seasonal and interannual variability of the marine bacterioplankton community throughout the water column over ten years. ISME J 9:563\u0026ndash;580\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang XX, Xu W, Liu Y, Cai MW, Luo ZH, Li M (2018) Metagenomics reveals microbial diversity and metabolic potentials of seawater and surface sediment from a hadal biosphere at the Yap Trench. Front Microbiol 9:2402\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen SF, Zhou YQ, Chen YR, Gu J (2018) fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34:884\u0026ndash;890\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBankevich A et al (2012) SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol 19:455\u0026ndash;477\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang DD et al (2019) MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ 7:e7359\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParks DH, Imelfort M, Skennerton CT, Hugenholtz P, Tyson GW (2015) CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res 25:1043\u0026ndash;1055\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChaumeil PA, Mussig AJ, Hugenholtz P, Parks DH (2022) GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics 38:5315\u0026ndash;5316\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParks DH, Chuvochina M, Chaumeil PA, Rinke C, Mussig AJ, Hugenholtz P (2020) A complete domain-to-species taxonomy for Bacteria and Archaea. Nat Biotechnol 38:1079\u0026ndash;1086\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQi YL et al (2024) Analysis of nearly 3000 archaeal genomes from terrestrial geothermal springs sheds light on interconnected biogeochemical processes. Nat Commun 15:4066\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKieft K, Zhou ZC, Anantharaman K (2020) VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences. Microbiome 8:90\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo JR et al (2021) VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome 9:37\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoux S et al (2019) Minimum information about an uncultivated virus genome (MIUViG). Nat Biotechnol 37:29\u0026ndash;37\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNayfach S, Camargo AP, Schulz F, Eloe-Fadrosh E, Roux S, Kyrpides NC (2021) CheckV assesses the quality and completeness of metagenome-assembled viral genomes. Nat Biotechnol 39:578\u0026ndash;585\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFu LM, Niu BF, Zhu ZW, Wu ST, Li WZ (2012) CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics 28:3150\u0026ndash;3152\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCamargo AP et al (2023) IMG/VR v4: an expanded database of uncultivated virus genomes within a framework of extensive functional, taxonomic, and ecological metadata. Nucleic Acids Res 51:D733\u0026ndash;D743\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGregory AC et al (2019) Marine DNA viral macro- and microdiversity from pole to pole. Cell 177:1109\u0026ndash;1123\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu SF et al (2021) DeePhage: distinguishing virulent and temperate phage-derived sequences in metavirome data with a deep learning approach. Gigascience 10:giab056\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShang JY, Jiang JZ, Sun YN (2021) Bacteriophage classification for assembled contigs using graph convolutional network. Bioinformatics 37:I25\u0026ndash;I33\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHyatt D, Chen GL, Locascio PF, Land ML, Larimer FW, Hauser LJ (2010) Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 11:119\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuchfink B, Xie C, Huson DH (2015) Fast and sensitive protein alignment using DIANOND. Nat Methods 12:59\u0026ndash;60\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuerta-Cepas J et al (2019) eggNOG 5.0: a hierarchical, functionally and phylogenetically annotated orthology resource based on 5090 organisms and 2502 viruses. Nucleic Acids Res 47:D309\u0026ndash;D314\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAramaki T et al (2020) KofamKOALA: KEGG Ortholog assignment based on profile HMM and adaptive score threshold. Bioinformatics 36:2251\u0026ndash;2252\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShaffer M et al (2020) DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res 48:8883\u0026ndash;8900\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMirdita M, Sch\u0026uuml;tze K, Moriwaki Y, Heo L, Ovchinnikov S, Steinegger M (2022) ColabFold: making protein folding accessible to all. Nat Methods 19:679\u0026ndash;682\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Kempen M et al (2024) Fast and accurate protein structure search with Foldseek. Nat Biotechnol 42:243\u0026ndash;246\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoux S et al (2023) iPHoP: An integrated machine learning framework to maximize host prediction for metagenome-derived viruses of archaea and bacteria. PLoS Biol 21:e3002083\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAroney STN, Newell RJP, Nissen JN, Camargo AP, Tyson GW, Woodcroft B (2025) CoverM: read alignment statistics for metagenomics. Bioinformatics 41:btaf147\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangdon WB (2015) Performance of genetic programming optimised Bowtie2 on genome comparison and analytic testing (GCAT) benchmarks. BioData Min 8:1\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlm MR, Crits-Christoph A, Bouma-Gregson K, Firek BA, Morowitz MJ, Banfield JF (2021) inStrain profiles population microdiversity from metagenomic data and sensitively detects shared microbial strains. Nat Biotechnol 39:727\u0026ndash;736\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDixon P (2003) VEGAN, a package of R functions for community ecology. J Veg Sci 14:927\u0026ndash;930\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8682536/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8682536/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eViruses influence microbial mortality and carbon flux in the ocean, yet how water-column depth shapes their ecology and evolution from surface waters to the hadal zone remains unexplored. To address this gap, we profiled a continuous 2\u0026ndash;6,000 m transect in the Yap Trench in the Pacific Ocean, amongst the world\u0026rsquo;s deepest ocean trenches using depth-resolved metagenomics. We reconstructed a total of 8,520 viral operational taxonomic units and observed the viral communities to be strongly stratified, with diversity peaking in the mesopelagic. Virulent \u003cem\u003eKyanoviridae\u003c/em\u003e dominated upper layers, while temperate \u003cem\u003ePeduoviridae\u003c/em\u003e increased and stabilized with depth, indicating a shift from \"kill the winner\" to \"piggyback the winner\" dynamics. Auxiliary metabolic genes mirrored this partitioning but were less defining than lifestyle transitions. Viruses infecting \u003cem\u003eGammaproteobacteria\u003c/em\u003e and \u003cem\u003eFirmicutes\u003c/em\u003e encoded cell motility and stress-response genes, suggesting adaptation to energy limitation. Evolutionary dynamics were depth-dependent: mutation rates declined with depth, while positive selection intensified in key viral genes. Overall, our results identify ocean depth as a central axis shaping viral community dynamics and evolutionary trajectories in marine ecosystems.\u003c/p\u003e","manuscriptTitle":"Depth-governed ecological and evolutionary partitioning of ocean trench viromes","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-12 17:20:18","doi":"10.21203/rs.3.rs-8682536/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"8e460b50-715b-4e8c-91f7-8ffc5f1e2c7e","owner":[],"postedDate":"February 12th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":62188588,"name":"Biological sciences/Ecology/Biooceanography/Microbial biooceanography"},{"id":62188589,"name":"Earth and environmental sciences/Ecology/Microbial ecology"},{"id":62188590,"name":"Biological sciences/Microbiology/Bacteriophages"},{"id":62188591,"name":"Biological sciences/Microbiology/Environmental microbiology/Water microbiology"}],"tags":[],"updatedAt":"2026-02-12T17:20:18+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-12 17:20:18","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8682536","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8682536","identity":"rs-8682536","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}