Bacteria and phage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks.

Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Microbiome . 2026 Mar 4;14:113. doi: 10.1186/s40168-026-02368-y Search in PMC Search in PubMed View in NLM Catalog Add to search Bacteria and phage consortia modulate cecal SCFA production and host metabolism to enhance feed efficiency in ducks Xia Wang Xia Wang 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Xia Wang 1, # , Shujie Tian Shujie Tian 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Shujie Tian 1, # , Yunsheng Zhang Yunsheng Zhang 2 Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193 China Find articles by Yunsheng Zhang 2, # , Li Yang Li Yang 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Li Yang 1 , Di Hu Di Hu 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Di Hu 1 , Zezhong Wang Zezhong Wang 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Zezhong Wang 1 , Xueqin Yang Xueqin Yang 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Xueqin Yang 1 , Shaofei Li Shaofei Li 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China Find articles by Shaofei Li 1 , Jie Wei Jie Wei 2 Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193 China Find articles by Jie Wei 2 , Wei Zhou Wei Zhou 2 Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193 China Find articles by Wei Zhou 2 , Shuaiqin Wang Shuaiqin Wang 2 Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193 China Find articles by Shuaiqin Wang 2 , Li Deng Li Deng 3 Prevention of Microbial Diseases, TUM School of Life Sciences; Central Institute of Infection Prevention (ZIP), Technical University of Munich, Freising, 85354 Germany 4 Research Unit of Microbial Diseases Prevention, Institute of Virology, Helmholtz Centre Munich-German Research Centre for Environmental Health, Neuherberg, 85764 Germany Find articles by Li Deng 3, 4 , Fuyong Li Fuyong Li 5 Department of Animal Science and Technology, College of Animal Sciences, Zhejiang University, Hangzhou, 310058 China 6 Ministry of Education Key Laboratory of Molecular Animal Nutrition, Zhejiang University, Hangzhou, 310058 China Find articles by Fuyong Li 5, 6, ✉ , Shuisheng Hou Shuisheng Hou 2 Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193 China Find articles by Shuisheng Hou 2, ✉ , Pengying Li Pengying Li 7 College of Landscape Architecture and Art, Northwest A&F University, Yangling, 712100 China Find articles by Pengying Li 7, ✉ , Jinlong Ru Jinlong Ru 3 Prevention of Microbial Diseases, TUM School of Life Sciences; Central Institute of Infection Prevention (ZIP), Technical University of Munich, Freising, 85354 Germany 4 Research Unit of Microbial Diseases Prevention, Institute of Virology, Helmholtz Centre Munich-German Research Centre for Environmental Health, Neuherberg, 85764 Germany Find articles by Jinlong Ru 3, 4, ✉ Author information Article notes Copyright and License information 1 College of Animal Science and Technology, Northwest A&F University, Yangling, 712100 Shaanxi China 2 Institute of Animal Sciences, Chinese Academy of Agricultural Sciences, Beijing, 100193 China 3 Prevention of Microbial Diseases, TUM School of Life Sciences; Central Institute of Infection Prevention (ZIP), Technical University of Munich, Freising, 85354 Germany 4 Research Unit of Microbial Diseases Prevention, Institute of Virology, Helmholtz Centre Munich-German Research Centre for Environmental Health, Neuherberg, 85764 Germany 5 Department of Animal Science and Technology, College of Animal Sciences, Zhejiang University, Hangzhou, 310058 China 6 Ministry of Education Key Laboratory of Molecular Animal Nutrition, Zhejiang University, Hangzhou, 310058 China 7 College of Landscape Architecture and Art, Northwest A&F University, Yangling, 712100 China ✉ Corresponding author. # Contributed equally. Received 2025 Aug 12; Accepted 2026 Feb 3; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . PMC Copyright notice PMCID: PMC13069815  PMID: 41782011 Abstract Background The gut microbiota influences poultry health, nutrition, feed efficiency (FE), and overall productivity. However, the relationship between gut microbes, including bacteria and phages, and FE in ducks remains underexplored. To address this, we integrated cecal 16S amplicon, metagenome, microbiota-derived short-chain fatty acids (SCFAs) profiling, liver transcriptome, and serum metabolome data to illustrate the contribution of the gut microbiome (bacteria and viruses) to duck FE. Results We reconstructed viral genomes and prokaryotic metagenome-assembled genomes (MAGs) and annotated their genes using comprehensive databases. Prokaryotic hosts of viruses were also predicted to understand virus-host dynamics within the gut ecosystem. Our results revealed that high-FE ducks have higher concentration of propionate and butyrate in cecum compared with low-FE ducks. The metagenome sequencing revealed distinct cecal microbiota profiles between two groups, with increased relative abundance of representative SCFA producers, especially Paraprevotella sp905215575 and Bacteroides sp944322345 , and enhanced SCFA-biosynthesis pathways in high-FE ducks. Virome genome assembly identified two phages encoding auxiliary metabolic genes (AMGs) involved in pyruvate metabolism, enhancing nutrient availability for host bacteria to produce SCFAs (e.g., temperate phage-encoded pyruvate phosphate dikinase) or exploiting host central metabolic pathways for viral replication (e.g., lytic phage-encoded formate C-acetyltransferase). Furthermore, these representative SCFA-producing bacteria and phage consortia were associated with serum metabolites (including L-histidine and 4-hydroxydecanedioylcarnitine) linked to duck FE. Conclusion Collectively, these findings provide novel insights into the gut microbial factors regulating FE in ducks, offering potential strategies to optimize poultry nutrition and productivity. Your browser is not supporting the HTML5 <video> element. You may download that video as a file and play it with a player of you choice. Download video stream . Download video file (46.2MB, mp4) Open in a new tab Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s40168-026-02368-y. Introduction Efficient feed utilization is critical for sustainable and profitable poultry production. Feed costs account for 60–70% of total production expenses in commercial poultry operations, making feed efficiency (FE) a primary determinant of economic viability [ 1 ]. In the context of a growing global demand for animal protein, enhancing FE in poultry species like chickens and ducks holds significant economic and environmental importance through lower feed waste and nitrogen excretion. In non-ruminant animals like poultry and pigs, short-chain fatty acids (SCFAs), primarily produced through microbial fermentation of fiber in the cecum or colon, can supply 60–70% of the energy requirements of intestinal epithelial cells [ 2 ]. This process not only sustains normal intestinal epithelial function but also plays a vital role in overall gut health and metabolic regulation. Beyond energy provision, SCFAs (such as acetate, propionate, and butyrate) function as signaling molecules by activating G protein-coupled receptors (GPCRs), including FFAR2 (GPR43), FFAR3 (GPR41), and HCAR2 (GPR109A). Activation of these receptors modulates energy expenditure, preadipocyte differentiation, and appetite control [ 3 , 4 ]. Given these functions, modulating the gut microbiota to enhance SCFA production represents a promising strategy for improving feed utilization. For example, probiotic supplementation promotes a faster shift toward a mature, stable microbiota composition, increasing average daily gain (ADG) by 6.2% while reducing FCR by 4.5% compared to antibiotic-free controls [ 5 ]. Similarly, microbiota-accessible fibers boost populations of SCFA-producing bacteria (e.g., Bacteroidetes ) by up to 40%, enhancing energy harvest and intestinal barrier function in laying hens [ 6 ]. While bacterial contributions to FE through SCFA production are well-documented [ 7 – 9 ], the specific roles of phages in modulating these bacterial functions and their combined impact on duck feed efficiency remain uncharacterized, preventing the development of comprehensive microbiome-based strategies to optimize feed utilization. The gut virome, dominated by phages (viruses infecting bacteria), represents one of the densest viral ecosystems in nature, with estimates of more than 10 8 to 10 10 virus-like particles per gram of intestinal content [ 10 , 11 ]. These phages exert profound influence over microbial community dynamics and metabolic processes through predator–prey interactions, horizontal gene transfer, and the regulation of bacterial lysogenic-to-lytic life cycles. For instance, phages can reshape bacterial diversity by selectively lysing dominant taxa like Bacteroides or Bacillota , thereby altering niche competition and nutrient flow [ 12 ]. Phage-encoded auxiliary metabolic genes (AMGs) further modulate bacterial metabolism, such as enhancing carbohydrate utilization in Bacteroides [ 13 ] and Oscillospiraceae [ 14 ]. Such phage activities can cascade to reshape microbial functional outputs, such as bile acid metabolism and SCFA production, with indirect implications for host physiology. Despite their ubiquity and functional significance, the role of gut viruses in shaping host phenotypes remains understudied outside human disease contexts (e.g., inflammatory bowel disease, where phage dysbiosis alters bacterial consortia and immunity [ 15 ]). In livestock, emerging evidence suggests agricultural relevance: rumen phages in cattle may regulate host nutrient utilization via AMGs and virus-host linkages, potentially influencing FE [ 12 , 16 ]. These findings position the virome as a hidden effector bridging microbial ecology and host biology. However, data on viral-bacterial interactions and their association with FE in ducks are limited, and specific microbial taxa or functional genes linked to duck FE remain unidentified. Elucidating these interactions is essential for developing microbiome-based strategies to enhance FE in ducks. This study characterized gut microbial communities, including bacteria and phages, in ducks stratified by FE. We hypothesized that high-FE ducks harbor distinct gut microbiota compositions and functional capabilities compared with low-FE ducks. We further aimed to resolve virus-host dynamics and assess their impact on bacterial community structure and function. By integrating 16S amplicon, metagenome, SCFAs, liver transcriptome, and serum metabolome data, we provide a comprehensive assessment of the duck gut microbiome’s association with FE. Our findings may inform microbiome-targeted strategies (e.g., probiotics, dietary interventions) to optimize feed utilization in duck production. Results FE traits and SCFAs in high-FE (ZF) and low-FE (CF) ducks Two Pekin duck populations with divergent FE were used: a high-FE group (ZF, n = 150) and a low-FE group (CF, n = 150) (see the “Methods” section for details). The statistics of average daily feed intake (ADFI), average daily gain (ADG), feed conversion ratio (FCR), and fatness traits between the two duck populations are shown in Fig. 1 A, B and Table S1 . Compared to ZF ducks, CF ducks exhibited 5.45% higher ADFI ( p < 0.0001), 4.59% lower ADG ( p < 0.001), 9.45% higher FCR ( p < 0.0001), 55.83% greater skin fat weight (SFW) ( p < 0.001), and 26.64% greater abdominal fat weight (AFW) ( p < 0.001). These results demonstrate significantly superior FE and reduced adiposity in the ZF population relative to CF ducks. Fig. 1. Open in a new tab FE traits and SCFAs in two groups. A and B Average daily gain (ADG) and feed conversion ratio (FCR) of CF and ZF ducks. C and D Absolute concentrations of propionate and butyrate in two groups. * p < 0.05, *** p < 0.001, **** p < 0.0001 SCFAs derived from gut microbiota have been proven to be associated with animal FE previously [ 8 , 9 ]. To compare cecal SCFAs in two groups, we analyzed 10 SCFAs from a randomly selected subset of both groups (20 CF, 16 ZF), which also showed a significant difference in FCR and ADG (Fig. S1 A, B), using targeted metabolomics. We found that ZF ducks exhibited higher propionate and butyrate concentrations in cecal content compared with CF ducks ( p < 0.05), while the others had no significant difference (Fig. 1 C, D, and Fig. S1 C–J). These results combined suggest that higher propionate and butyrate are associated with improved FE in ducks. Divergence of cecal bacterial diversity, composition, and function in high- and low-FE ducks SCFAs in duck are primarily derived from microbial fermentation of fiber in the cecum; we therefore first investigated cecal bacterial composition and interactions in CF and ZF groups using 16S rRNA gene amplicon sequencing. The data revealed a conserved phylum-level structure across both groups, dominated by Bacteroidetes followed by Bacillota_A and Bacillota_C (Fig. 2 A and Fig. S2A–F). While observed ASV richness was comparable between groups, ZF ducks exhibited significantly elevated α-diversity versus CF ducks as measured by Chao1, Pielou, and Shannon indices (Fig. 2 B). The co-occurrence networks of the two groups have low adjusted rand index (ARI) (0.136) [ 17 ], indicating that the network’s structural and topological characteristics varied for each group. Detailed information on network topology is shown in Fig. S2G and Tables S2–S3. Fig. 2. Open in a new tab Taxonomic compositions, diversity, and co-occurrence networks of gut bacterial communities revealed by 16S rRNA gene amplicon sequencing. A The average composition of top 15 bacterial communities at phylum level in CF and ZF duck populations. * p < 0.05, ** p < 0.01. B α-diversity (Chao1, Pielou, Richness, and Shannon indices) in two groups. C Differential abundance of bacterial species between two groups ( p < 0.05). Values represent natural log fold changes from ANCOM-BC analysis. The color gradient indicates the direction and magnitude of change: blue to white represent decreases, while white to red represent increases in ZF compared with CF. D Pearson correlation and linear regression analyses between bacterial species and feed conversion ratio (FCR). E Correlation and linear regression analyses between bacterial species and SCFAs (butyrate and propionate). The correlation coefficient r and the corresponding p -value are displayed in each plot. The regression line (gray) is shown with its 95% confidence interval (shaded area). F Non-metric multidimensional scaling (NMDS) analysis of the Bray–Curtis distances based on gut microbial species. Ellipsoids represent a 95% confidence interval surrounding each group. G Relative abundance (total sum scaling) of aceE ( K00163 ; pyruvate dehydrogenase) in two groups. H Bar plot showing the p -values (-log10 transformed) of differentially abundant KOs in modules. Red, ZF-enriched KOs; blue, CF-enriched KOs. I Metabolic pathways involved in SCFAs biosynthesis and methanogenesis To further identify the specific intestinal microbial species that contribute to FE, we analyzed cecal microbiome from the subset of both groups (20 CF, 16 ZF) using metagenomics. Prokaryotic genome reconstruction from metagenomic data yielded 1820 MAGs through single-sample assembly (mean contigs per MAG 482). Quality filtering (completeness ≥ 50%, contamination < 10%) retained 943 medium-to-high-quality MAGs. Taxonomic classification assigned these MAGs to 16 phyla, with 890 MAGs (94.4%) resolving to 121 genera and 620 MAGs to 144 known species. Notably, 321 MAGs (34.0%) lacked matches to reference genomes, representing potentially novel species. We performed the analysis of the composition of microbiomes with bias correction (ANCOM-BC) and identified 13 species significantly different in terms of abundance between two groups (adjusted p < 0.05) (Fig. 2 C). Specifically, gut bacterial communities in ZF ducks were mainly enriched in these species, such as UMGS263 sp949289045 , Paraprevotella stercoravium , Paraprevotella sp905215575 , Bacteroides pullorum , and Bacteroides sp944322345 , while CF animals were more abundant in Alistipes finegoldii , Desulfovibrio sp944324525 , Fusobacterium A mortiferum , and Avelusimicrobium sp944324635 . We then performed correlation analyses between these differentially abundant bacterial species and FCR/ADG among individuals. Multiple species, including UMGS263 sp949289045 , Paraprevotella sp905215575 , and Bacteroides sp944322345 , were negatively correlated with FCR ( r = − 0.43 ~ − 0.37, p < 0.05) but positively correlated with ADG (Fig. 2 D and Fig. S2H). We further analyzed the correlation of these differential bacterial species with propionate or butyrate; the results revealed multiple differential bacteria, such as Paraprevotella sp905215575 and Bacteroides sp944322345 , were positively correlated with propionate or butyrate ( r = 0.36 ~ 0.55, p < 0.05), while a few species, including Desulfovibrio sp944324525 and Alistipes finegoldii , displayed a negative association with propionate (Fig. 2 E and Fig. S2H). To further compare the functional features of microbiota in two duck groups, we annotated the microbial functions of each metagenome via the KEGG database using a gene-centric approach [ 18 ] and quantified the abundance of each KEGG orthologous (KOs) in different samples. Non-metric multidimensional scaling (NMDS) plot showed an obvious difference in the gut microbial KOs between the CF and ZF groups (PERMANOVA, R 2 = 0.167, p < 0.001; Fig. 2 F). We identified 545 differentially abundant KOs ( p < 0.05), with 339 (62.2%) enriched in ZF ducks. These differential KOs were significantly enriched in multiple KEGG pathways including “carbon metabolism”, “sulfur metabolism”, “phosphonate and phosphinate metabolism”, “methane metabolism”, “citrate cycle (TCA cycle)”, “nucleotide metabolism”, and “pyruvate metabolism” (adjusted p < 0.1, Fig. S2I). As “pyruvate metabolism” is a central hub in the anaerobic glycolysis pathway of gut bacteria, driving the synthesis of short-chain fatty acids (SCFAs) like acetate, propionate, and butyrate, ZF microbiomes showed significant enrichment of aceE ( K00163 ; pyruvate dehydrogenase) within pyruvate oxidation modules (M00307), converting pyruvate to acetyl-CoA, the central precursor for SCFA biosynthesis (Fig. 2 G). This predicted enhanced SCFA production was consistent with our experimental measurement that ZF ducks exhibited higher propionate and butyrate concentrations in cecal content compared with CF ducks. Regarding carbohydrate metabolism and energy metabolism, KEGG pathway enrichment analysis (Fig. S2I) revealed that the differential KOs were significantly enriched in several pathways. Notably, the “methane metabolism” pathway (in which 9 of the 12 differential KOs were higher in CF) was predominantly enriched in the CF microbiome. Conversely, the “Citrate cycle (TCA cycle)” (5/6 KOs enriched in ZF), “Pentose and glucuronate interconversions” (6/6 KOs enriched in ZF), and “Phosphonate and phosphinate metabolism” (5/6 KOs enriched in ZF) pathways were significantly more abundant in the gut microbiome of the ZF group (Fig. 2 H, I). These enrichment patterns align with known energy-partitioning mechanisms: ZF’s enhanced pentose/glucuronate interconversions promote SCFA yield [ 19 , 20 ], while CF’s methane metabolism redirects carbon from SCFA production [ 21 ]. Gut virome characterization and host interactions Metagenomic assembly of 36 cecal samples yielded 4334 quality-filtered viral contigs (Fig. 3 A). Among these, 835 high-confidence contigs (completeness ≥ 50%) had a mean length of 41,983 bp (N50 = 50,390 bp). Clustering at 95% average nucleotide identity (ANI) and 85% alignment fraction (AF) identified 1986 viral operational taxonomic units (vOTUs), including 428 with ≥ 50% completeness. All vOTUs represented phages, predominantly within Caudoviricetes (tailed phages), Malgrandaviricetes (microviruses), Megaviricetes , and Arfiviricetes (Fig. 3 B). Protein-sharing clustering of high-quality vOTUs generated 392 viral clusters (VCs). Host prediction revealed a total of 1,984 virus-host linkages, comprising 808 lytic VC-host and 1176 temperate VC-host linkages. The hosts of temperate phages spanned across 11 phyla (Fig. 3 C), whereas those of lytic phages were associated with 12 phyla (Fig. 3 D). As expected, Bacteroidota (comprising 58.17% of lytic and 69.56% temperate viral hosts) and Bacillota_A (comprising 15.97% of lytic and 15.73% temperate viral hosts), highly predominant gut bacterial taxa, were the most common hosts of gut viruses at the phylum level. Bacteroides and Phocaeicola were the most prevalent hosts at the genus level (Fig. S3A, B). Among these VC-host linkages, VC2778, VC2779, and VC2563 exhibited the highest host diversity, each associated with 62, 54, and 49 distinct host species (Fig. S3C). Fig. 3. Open in a new tab The viral contig characteristics and virus-host associations from 36 cecal metagenomes. A The distribution of the estimated completeness of 4,334 viral contigs. B The classification information of the class level of 1,986 vOTUs. C , D Number of temperate and virulent viral clusters (VCs) with > 50% completeness for each predicted host at the phylum level Virome divergence and metabolic modulation in high- and low-FE Ducks To compare gut viral communities between ZF and CF ducks, we examined viral α- and β-diversity of two groups. The results showed that the two groups exhibited similar α-diversity profiles except for the Shannon index at the VC level (Fig. 4 A), while β-diversity analysis (NMDS, Bray–Curtis) revealed distinct clustering of individuals from the two groups (PERMANOVA, R 2 = 0.096, p < 0.001; Fig. 4 B). Fig. 4. Open in a new tab Viral diversity and their relationship with FE related traits. A Viral diversity (Chao1, Pielou, Richness, Shannon index) between two groups. B The NMDS plot of cecal viral communities between two groups based on Bray–Curtis dissimilarity. C Spearman’s rank correlation between FE traits and differential VCs, with color gradients representing positive (red) and negative (blue) correlation coefficients ( p < 0.05). D Spearman’s rank correlation between differential VCs and differential bacteria (CF-enriched bacteria are colored red). *** p < 0.001, ** p < 0.01, * p < 0.05 To further investigate gut viral signatures associated with FE, we performed ANCOM-BC analysis and identified that 32 VCs were significantly decreased, while 19 VCs were increased in CF group compared with ZF group (Fig. S3D). Then we did the correlation analysis between these differential VCs and FE-related phenotypes. Twelve viruses (5 positive and 7 negative) and 6 viruses (4 positive and 2 negative) displayed significant correlations with FCR and ADG, respectively (Spearman, | r |≥ 0.3, p < 0.05) (Fig. 4 C). For example, VC2859, VC2785, and VC0894 showed consistent positive ADG or negative FCR correlations. As phages may link to phenotype by modulating the composition and dynamics of bacterial communities, we predicted interactions of these FE-associated VCs (either co-occurrence or mutual exclusion) with differentially abundant bacteria (Fig. 4 D). Spearman correlation analysis revealed distinct phage-bacteria association patterns. One group of VCs, including VC2785, VC0894, and VC2518, displayed significant positive correlations (co-occurrence) with high-FE favored bacteria (e.g., Paraprevotella sp905215575 , Bacteroides sp944322345 , UMGS263 sp949289045 ) ( r = 0.41 ~ 0.67, p < 0.05), while exhibiting significant negative correlations (mutual exclusion) with low-FE favored bacteria (e.g., Alistipes finegoldii , Desulfovibrio sp944324525 ) ( r = − 0.70 ~ − 0.47, p < 0.05). Conversely, a second group of VCs (e.g., VC2862, VC2755, VC2722, VC1483) displayed the opposite association pattern, positively correlating with low-FE favored bacteria and negatively correlating with high-FE favored bacteria. Importantly, to distinguish ecological interactions from direct infection, we cross-referenced these correlations with our in silico host predictions. We found that key strong positive correlations, such as that between VC2785 and Paraprevotella , were concordant with predicted phage-host linkages, reinforcing the biological validity of these associations. Then we searched AMGs from viral contigs and found 109 AMGs that were involved in a wide range of metabolic processes, including the metabolism of nucleotides, carbohydrates, amino acids, and sulfur (Fig. 5 A). The most prevalent AMG was DNMT1, a DNA (cytosine-5)–methyltransferase that prevents viruses from being recognized and cleaved by the restriction-modification systems of their hosts [ 22 , 23 ], which were detected in more than 40 high-quality viral contigs. CysH was the second highly abundant AMG, which encodes a phosphoadenosine phosphosulfate reductase involved in the synthesis of sulfite from sulfate. Additionally, another abundant AMG, metK encoding S-adenosylmethionine (SAM) synthetase, was important for genomic DNA methylation and cell division. Two AMGs, NAMPT and pncA, were implicated in NAD + metabolism, which is critical for maintaining energy homeostasis. CobS and cobT were two key AMGs involved in the synthesis of vitamin B12. Fig. 5. Open in a new tab Auxiliary metabolic genes (AMGs) carried by duck cecal viruses. A A bar plot showing the occurrence (log10) of AMGs identified in the temperate (green) and virulent (yellow) viruses. B The relative abundance of two AMGs (PPDK and pflD) in two groups. *** p < 0.001, * p < 0.05. C Genome synteny maps of two viruses (ctg4378 is identified in this study) and one bacterium containing the PPDK gene. D Genome synteny maps of two viruses (ctg2056 is identified in this study) and one bacterium containing the pflD gene. Homologous genes from different organisms are filled in the same color and connected using the shading links. E Predicted tertiary structure and function of the phage encoded PPDK protein. The phage-encoded PPDK structure is superimposed onto the reference structure of PPDK from Clostridium symbiosum (UniProt: P22983 ; colored silver) based on Foldseek search [ 24 ]. PPDK (EC 2.7.9.1) catalyzes the conversion of AMP, phosphoenolpyruvate, and diphosphate to ATP, pyruvate, and inorganic phosphate. F Predicted tertiary structure and function of the phage encoded pflD protein. The phage-encoded pflD structure is superimposed onto the reference structure of pflD from Enterococcus faecium (UniProt: A0A132P2N7; colored silver) based on Foldseek search. pflD (EC 2.3.1.54) catalyzes the conversion of pyruvate and CoA to formate and acetyl-CoA We subsequently conducted a differential analysis of AMGs between CF and ZF groups and found 16 AMGs showing significant differential abundance. Notably, two differential AMGs, pyruvate phosphate dikinase (PPDK) and formate C-acetyltransferase (pflD), involved in pyruvate metabolism were identified in viral contigs (Fig. 5 B). This PPDK shared 83% amino acid identity with PPDK in Prevotella sp. E2-28 (GenBank: CP091788 ) (Fig. 5 C), while the pflD was 52% identical to a pflD in Catenibacterium mitsuokai strain DSM 15897 (GenBank: CP102271 ) (Fig. 5 D). Specifically, in the anaerobic cecal environment, temperate virus (VC2785)-encoded PPDK, catalyzing the reversible conversion of phosphoenolpyruvate (PEP) to pyruvate and ATP [ 25 ], was higher abundant in the ZF group compared with the CF group, which may augment the fermentative capacity of host bacteria (e.g., Bacteroides and Paraprevotella ) in the ZF group (Fig. 5 E). This is in line with the abovementioned observation that VC2785 displayed a positive correlation with both ADG (Fig. 4 D). While lytic virus (VC0717)-encoded pflD, converting pyruvate to acetyl-CoA and formate, was higher abundant in the CF group compared with the ZF group, which may accelerate viral replication by hijacking the host’s metabolic machinery in the CF group (Fig. 5 F). Moreover, its host, Prevotella sp015074785 , belonging to the Prevotellaceae family, significantly contributes to the production of SCFAs, which may induce lytic viral replication and lysis of cells [ 26 , 27 ]. These results together suggest that gut viruses could potentially optimize nutrient availability for their hosts or exploit host central metabolic pathways for their own replication, underscoring their influence on host physiology including FE. Serum metabolome profiling reveals systemic metabolic divergence between two groups Serum metabolomics identified 211 differentially abundant metabolites between CF and ZF ducks (|log 2 FC|≥ 1, p < 0.05), with 175 upregulated and 36 downregulated in CF compared with ZF (Fig. S4). According to the annotation of the Human Metabolome Database (HMDB), these differential metabolites were mainly categorized into lipids and lipid-like molecules (38.39%), organic acids and derivatives (10.43%), organoheterocyclic compounds (9.95%), and organic oxygen compounds (5.21%) (Fig. 6 A). Metabolite pathway enrichment analysis indicated that these differential metabolites were mainly enriched in 13 pathways (Fig. 6 B). Among them, 3 pathways including “pyrimidine metabolism”, “butanoate metabolism”, and “pantothenate and CoA biosynthesis” showed concordant enrichment in liver transcriptomes (Fig. 6 C). For example, in butanoate metabolism, we observed higher abundance of succinic semialdehyde in the CF group compared with the ZF group (log 2 FC = 1.37, p < 0.05); this may stem from downregulation of the succinic semialdehyde dehydrogenase (ALDH5A1) gene (log 2 FC = − 0.75, p < 0.01), a mitochondrial homotetramer protein that catalyzes the NAD + -dependent conversion of succinic semialdehyde into succinate, which can be further fermented to butanoate (Fig. 6 D, E). In functional enrichment analysis of DEGs, we also observed multiple pathways related to lipid metabolism, such as “insulin resistance”, “PPAR signaling pathway”, and “glycerophospholipid metabolism”, explaining the predominant lipid metabolite differences observed between the two groups (Fig. 6 C). Fig. 6. Open in a new tab Serum metabolome and liver transcriptome profiles reveal systemic metabolic divergence between the two groups. A Classification and composition of differentially abundant serum metabolites between two groups. B Functional enrichment analysis of differentially abundant metabolites and the resulting p -values were adjusted for multiple testing using the Benjamini–Hochberg FDR method. C KEGG pathway enrichment analysis of differentially expressed genes (DEGs) in the liver between two groups and the resulting p -values were adjusted for multiple testing using the Benjamini–Hochberg FDR method. D Relative abundance of succinic semialdehyde in two groups. E Expression level of succinic semialdehyde dehydrogenase (ALDH5A1) in two groups. ** p < 0.01, * p < 0.05 Gut microbiome-serum metabolome associations with feed efficiency Among differential serum metabolites, 36 metabolites were significantly correlated with FE traits (| r |≥ 0.3, p < 0.05), with most of them positively correlated with FCR (21/36) (Fig. 7 A). For example, L-histidine and 4-hydroxydecanedioylcarnitine were negatively correlated with ADG, while positively associated with FCR. To detect the potential effect of differentially abundant bacteria on host FE-associated metabolites, spearman’s rank correlations revealed a total of 67 strong correlations (| r |≥ 0.5, p < 0.05; Fig. 7 B). Among these correlations, most of them were negative between differential bacteria and FE-associated metabolites (e.g., Paraprevotella sp905215575 vs. L-histidine: r = − 0.67), while these bacteria were positively correlated with two metabolites (e.g., UMGS263 sp949289045 vs. CDP-DG (a-25:0/20:4(6E,8Z,11Z,14Z) + = O(5)): r = 0.48). Spearman’s rank correlations between differential viruses and FE-associated metabolites were also performed, most FE-associated metabolites showed positive correlations with 12 VCs (e.g., VC2540 vs. 4-hydroxydecanedioylcarnitine: r = 0.64) and negative correlations with 13 VCs (e.g., VC2785 vs. L-histidine: r = − 0.56), respectively, while the correlation of CDP-DG(a-25:0/20:4(6E,8Z,11Z,14Z) + = O(5)) and (2R)-O-Phospho-3-sulfolactate with VCs showed the reversed pattern (Fig. 7 C). Fig. 7. Open in a new tab Correlations between gut microbiome and serum metabolome. A Spearman’s rank correlations between differential serum metabolites and FE traits. B Spearman’s rank correlations between FE-associated metabolites and differentially abundant bacteria. C Spearman’s rank correlations between FE-associated metabolites and differential viruses. Metabolites are clustered by their biochemical classifications. *** p < 0.001, ** p < 0.01, * p < 0.05 Discussion In the past decade, gut microbiome and its associated roles have attracted increasing attention. Although several studies have investigated the bacterial contributions to FE of poultry using 16S rRNA sequencing [ 8 , 28 ], the role of gut virome remains poorly characterized. Our multi-omics integration (metagenomics, transcriptomics, metabolomics) reveals associations between bacterial and viral components of the duck gut microbiome and FE-associated phenotypes. We identified that ZF ducks have higher FE traits and cecal SCFAs (e.g., propionate and butyrate) concentration compared with CF group. Due to pivotal roles of SCFAs in maintaining energy homeostasis and promoting animal health through multifaceted mechanisms, including maintenance of intestinal barrier integrity and improvement of nutrient absorption [ 29 ], thus animals producing more SCFAs usually may have high FE. Previous studies showed that butyrate supplementation improved FE in chicken and duck [ 30 , 31 ]. In a sheep study, higher concentration of propionate was also detected in rumen of high FE group [ 9 ]. In duck, SCFAs are chiefly produced in the cecum; thus, the difference of SCFAs between two groups may arise from variation in cecal bacterial diversity, composition, and function. We found that cecal bacteria in high-FE (ZF) and low-FE (CF) ducks shared core phyla ( Bacteroidetes and Bacillota ) but exhibited divergent diversity and network architectures. ZF microbiomes displayed higher modularity (Fig. S2G, H), indicative of more persistent and stable community-level and within-community interactions [ 32 , 33 ]. Besides network topology, the relative abundances of bacteria at various taxonomic levels in the two groups also exhibit differences; most of the differential species are enriched in the ZF group. Crucially, these ZF-enriched bacteria (e.g., Paraprevotella sp905215575 and Bacteroides sp944322345 ) potentially drove efficient pyruvate flux toward SCFA production rather than methanogenesis (Fig. 4 E, F). This is consistent with the observation of elevated cecal propionate/butyrate in ZF ducks. It has been reported that Paraprevotella was more enriched in the leaner pigs [ 34 ], which usually have high FE traits [ 8 , 35 ]. Conversely, CF-enriched Desulfovibrio sp944324525 and Alistipes finegoldii correlated negatively with SCFAs (Fig. S2H), aligning with inefficient energy harvesting; this was also observed in chicken gut [ 36 ]. As the “dark matter” of the gut microbiome, phages (predominantly Caudoviricetes ) can alter the bacterial community structure by lysing certain bacteria or through lysogeny, thereby affecting animal phenotypes. There was a significant difference in β diversity of viruses, suggesting a distinct viral community structure in two groups. In addition, multiple differentially abundant viruses showed a significant correlation with FE traits in ducks. From the predicted phage–host linkages, we found that some FE-correlated phages could target SCFA producers, such as Paraprevotella sp905215575 and Bacteroides sp944322345 , which are favorable for high FE. These results position phages as key phenotype mediators, which is supported by our human gut findings [ 14 ]. During the phage infection of bacteria, AMGs carried by phages play crucial roles in optimizing the host’s metabolism and nutrient acquisition, which affect several important ecological processes, including methane metabolism in different habitats [ 37 ], polysaccharide digestion in the ruminants [ 12 ], and SCFAs metabolism in the human gut [ 14 ]. However, the contribution of viral AMG in poultry gut has not been reported yet. Our results identified a total of 109 AMGs in duck cecum, which participated multiple processes, such as antiviral defense (e.g., DNMT1), sulfur metabolism (e.g., cysH and metK), energy homeostasis (e.g., NAMPT and pncA), and vitamin biosynthesis (cobS and cobT), suggesting that the gut virome can also affect feed digestibility in duck. We identified a total of 16 differentially abundant AMGs between two groups, with two AMGs, PPDK and pflD, involved in pyruvate metabolism process. In anaerobic environments like the cecum, microbes rely on fermentation pathways. PPDK enables the regeneration of PEP, which serves as a key intermediate for substrate-level phosphorylation, a primary ATP-generating mechanism under anaerobic conditions [ 25 ]. Given the requisite integration of temperate phages into the host genome, the temperate phages-encoded PPDK enriched in ZF potentially boosted pyruvate metabolism and SCFAs production of host bacteria, thereby conferring communal benefits and, in turn, reciprocal advantages to the viruses themselves [ 38 ]. While the lytic phage-encoded pflD enriched in CF could hijack host resources to produce more acetyl-CoA (SCFAs), which may promote viral infection and host lysis [ 39 ]. However, the underlying mechanisms remain to be explored and need further experimental validation. Gut microbiota produce a variety of molecules, some of which are taken up into the bloodstream and affect host health. Therefore, characterizing the interactions between gut microbiota and host plasma metabolites is also important for understanding the effects of the gut microbiota on animal phenotypes. Our results revealed that L-histidine and succinic semialdehyde were more abundant in serum of the CF group compared with the ZF group. Consistently, recent papers have reported that plasma L-histidine was downregulated in lambs with high ADG [ 40 , 41 ]. Besides, our RNA-seq data revealed reduced hepatic ALDH5A1 expression in CF ducks, which could impair succinic semialdehyde → butanoate conversion and lead to accumulation of succinic semialdehyde (Fig. 7 D). Intriguingly, succinic semialdehyde levels were found to be significantly positively correlated with insulin resistance and liver fat content [ 42 ], concordant with dysregulated lipid pathways, such as “insulin resistance”, “PPAR signaling pathway”, and “glycerophospholipid metabolism”, in the CF group. We reported that these serum FE-related metabolites were associated with duck cecal microorganisms, where most of them displayed strong negative associations with multiple differential bacterial and viral species between groups. The negative correlations of plasma L-histidine level with Bacteroides and Paraprevotella are in harmony with the results observed in piglets, rats, and fish [ 43 – 45 ]. This is probably due to the contribution of Bacteroides and Paraprevotella genera to L-histidine degradation [ 46 , 47 ]. Due to “virus-host” interactions, where viruses lyse host bacteria or boost their metabolism, gut viruses also display correlations with plasma metabolites. However, these links still require experimental validation in further study. Such information will provide evidence highlighting the possibility of animal blood metabolites through manipulating gut microbial functions. Conclusion Taken together, to elucidate the mechanisms by which gut microbiota affect FE of ducks, we proposed two potential convergent pathways: (1) SCFAs are the major end products from the fermentation of gut bacteria. Viruses directly target SCFA-producing bacteria, altering community function and output, thereby jointly impacting FE through collaborative interactions with bacteria; (2) gut bacteria and viruses interact to modulate host animal metabolism via blood circulation and thus influencing FE (Fig. 8 ). However, further studies are required to validate the actual transcriptional activity, function of phage-encoded AMGs, and causal relationships between representative SCFA-producing bacteria and phage consortia and host physiological outcomes. Fig. 8. Open in a new tab The potential mechanism by which bacteria and phage regulate cecal SCFA production and host metabolism to improve FE in ducks. Phages modulate bacteria to produce SCFAs through encoded AMGs, PPDK, and pflD Materials and methods Animals management Two Pekin duck populations with divergent FE were used: a high-FE group (ZF) and a low-FE group (CF). A total of 300 1-day-old male ducklings (150 per group) with similar initial body weights were housed individually in wire-floor cages (200 × 100 × 40 cm) equipped with nipple drinkers and tubular feeders at the National Changping Comprehensive Agricultural Engineering and Technology Research Center (Livestock and Poultry Sub-Center). Ducks were reared in two phases (days 1–14 and days 15–42) according to standardized practices. Both groups received identical commercial diets (Hope Feed Co., Ltd., Fangshan District, Beijing, China); diet composition and nutritional components matched those described in our previous study [ 48 ]. Feed efficiency measurement and sample collection Body weight (BW) was recorded at 22 and 42 days of age. Total feed intake (FI) was measured from 22 to 42 days. Average daily feed intake (ADFI), average daily gain (ADG), and feed conversion ratio (FCR) were calculated as follows: ADFI ( g / day ) = Total FI / 21 days 1 ADG ( g / day ) = ( BW 42 - BW 22 ) / 21 days 2 FCR = Total FI / ( BW 42 - BW 22 ) 3 Valid phenotypic records were obtained for 283 ducks. After a 12-h fast, ducks were humanely euthanized according to standard commercial protocols. Fatness traits were recorded, including skin fat weight (SFW), skin fat percentage (SFP), abdominal fat weight (AFW), and abdominal fat percentage (AFP), calculated as. SFP ( % ) = ( SFW / carcass weight ) × 100 4 AFP ( % ) = ( AFW / carcass weight ) × 100 5 DNA extraction, 16S amplicon and metagenomic sequencing Ducks with valid FCR records are euthanized by cervical dislocation and dissected. Cecal contents (including chyme and mucosa) were immediately collected, snap-frozen in liquid nitrogen, and stored at − 80 °C until further processing. Microbial DNA was extracted from cecal samples of 283 ducks using the QIAamp DNA Stool Mini Kit (QIAGEN, Hilden, Germany). The V3–V4 hypervariable region of the bacterial 16S rRNA gene was amplified using primers 338 F (5′-ACTCCTACGGGAGGCAGCA-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′). Amplified libraries were sequenced (2 × 250 bp) on an Illumina NovaSeq 6000 platform. A subset of 40 ducks (20 CF, 20 ZF) was randomly selected. Microbial DNA was extracted from cecal content using a CTAB-based method. DNA quality and quantity were assessed using a Qubit 3.0 Fluorometer (Invitrogen). Four ZF samples were excluded due to insufficient DNA yield, resulting in 36 samples (20 CF, 16 ZF) for library preparation. Libraries were constructed using the VAHTS Universal Plus DNA Library Prep Kit (Illumina, San Diego, CA, USA) and sequenced (2 × 150 bp) on an Illumina NovaSeq X platform following standard protocols. SCFA concentration measurement Cecal samples from the 36 ducks used for metagenome sequencing were also analyzed for the concentration of 10 SCFAs: acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid, valeric acid, hexanoic acid, heptanoic acid, nonanoic acid, and decanoic acid. Briefly, samples were thawed on ice. Approximately 0.5 g of cecal content was homogenized in 1 mL of deionized water for 4 min at 40 Hz, followed by ultrasonic oscillation for 15 min on ice. The homogenate was centrifuged at 5000 rpm for 20 min at 4 °C. A 0.8-mL aliquot of the supernatant was mixed with 0.1 mL of 50% (v/v) sulfuric acid (H 2 SO 4 ) and 0.8 mL of extraction solution containing 25 mg/L internal standard (methyl tert-butyl ether, MTBE). The mixture was vortexed vigorously for 10 min and centrifuged at 10,000 rpm for 15 min at 4 °C. The resulting supernatant was analyzed using a SHIMADZU GC2030-QP2020 NX gas chromatography–mass spectrometer (GC–MS). Absolute SCFA concentrations were determined using external calibration curves generated from standard solutions of each SCFA at varying concentrations. Determination and analysis of plasma metabolome Plasma metabolites were analyzed using ultra-high-performance liquid chromatography (UHPLC) coupled with tandem mass spectrometry (MS/MS). Chromatographic separation was performed on an ACQUITY UPLC® HSS T3 column (100 mm × 2.1 mm, 1.8 μm average particle size, Waters Corporation) using 0.1% (v/v) formic acid aqueous solution and acetonitrile as mobile phases. To mitigate analytical bias from instrumental variation, a pooled quality control (QC) sample was created by combining equal aliquots of all sample supernatants. This QC sample was injected at regular intervals throughout the analytical sequence. Raw MS data were acquired using MassLynx™ V4.2 software (Waters Corporation). Data processing—including peak detection, alignment, and integration—was performed in Progenesis QI (Nonlinear Dynamics). Metabolite annotation was conducted against the METLIN database ( https://metlin.scripps.edu ) within the Progenesis QI platform, using a mass accuracy threshold of ≤ 10 ppm and MS/MS spectral matching. Preparation of libraries for RNA-sequencing and data analysis Liver tissues were collected from three CF and three ZF ducks immediately post-euthanasia, snap-frozen in liquid nitrogen, and stored at − 80 °C. Total RNA was extracted using TRIzol™ Reagent (Thermo Fisher Scientific, Waltham, MA) following manufacturer protocols. RNA integrity was assessed by UV spectrophotometry (NanoDrop™ 2000; Thermo Fisher Scientific) and Electropherogram analysis (Agilent 2100 Bioanalyzer; Agilent Technologies, Santa Clara, CA, USA). Ribosomal RNA was depleted using the Illumina TruSeq Stranded Total RNA Library Prep Kit with Ribo-Zero Gold. Resulting mRNA was fragmented and reverse-transcribed into cDNA libraries using the same kit. Libraries were sequenced (2 × 150 bp) on an Illumina HiSeq X platform. Raw reads were quality-trimmed and adapter-filtered using fastp (v0.20.0) [ 49 ]. Processed reads were mapped to the Pekin duck reference genome (ZJU1.0) using STAR (version 2.7.9a) [ 50 ] with default parameters. Gene-level counts were generated with STAR’s –quantMode GeneCounts option and analyzed for differential expression using DESeq2 (v1.26.0) [ 51 ] in R/Bioconductor, with normalization via the median-of-ratios method. 16S amplicon sequencing data processing Raw sequencing reads underwent quality assessment using FastQC. Adapters and low-quality bases were trimmed using fastp (v0.20.0). Processed reads were denoised, error-corrected, and filtered for chimeras using the DADA2 plugin (v1.18) [ 52 ] in QIIME2 (v2020.6) [ 53 ], generating amplicon sequence variants (ASVs). ASVs were filtered to retain those with relative abundance > 0.0001% (10⁻ 6 ) and presence in ≥ 2 samples. Taxonomy was assigned against the Genome Taxonomy Database (GTDB r207) using DADA2’s native Bayesian classifier with default parameters. Alpha diversity (Shannon, Chao1, Pielou, and Richness) metrics were calculated in QIIME2. Phyloseq (v1.34.0) [ 54 ] was used for evaluation of sequencing depth and preparation of final ASV tables for downstream statistical analysis. Metagenome-assembled genomes (MAGs) recovery and annotation Raw metagenomic reads were preprocessed using fastp (v0.20.0) for adapter trimming and quality filtering. Host-derived reads were removed by alignment to the duck reference genome (ZJU1.0) using Bowtie2 (v2.4.2) [ 55 ]. De novo assembly of filtered reads was performed per sample using metaSPAdes (v3.15.0) [ 56 ] with default parameters. Assembled contigs were subsequently binned into MAGs using an integrative approach with MetaBAT2 (v2.17) [ 57 ], MaxBin2 (v2.2.7) [ 58 ], and CONCOCT (v1.1.0) [ 59 ]. Consensus bins were generated and refined using DAS Tool (v1.1.3) [ 60 ]. The quality of the refined MAGs was assessed using CheckM2 (v1.0.1) [ 61 ] with default parameters. Taxonomic classification was performed against GTDB r207 using GTDB-Tk (v2.1.0) [ 62 ]. Community-level functional annotations were performed using HUMAnN 3.0 (v3.6) pipeline with the clean reads as input [ 63 ]. Species-level differential abundance between FE groups was determined using ANCOM-BC [ 64 ]. Open reading frames (ORFs) were predicted from all contigs using Prodigal (v2.6.3) [ 65 ] in metagenomic mode. Proteins were annotated using eggNOG database with eggNOG-mapper v2 [ 66 ]. A non-redundant gene catalog was constructed with CD-HIT (v4.8.1) [ 67 ] at 95% nucleotide identity and 90% coverage. Gene abundances were quantified by aligning reads to the non-redundant gene catalog using Salmon (v1.10.2) [ 68 ] in alignment-based mode. Differential abundance of KEGG Orthologs (KOs) was tested using MaAsLin2 (v1.12.0) [ 69 ] with default parameters and significance was determined based on p -value < 0.05, followed by FDR correction for pathway enrichment analysis. Viral contig identification and annotation Putative viral contigs were first filtered from assembled metagenomic contigs using geNomad (v1.8.0) [ 70 ] with a relaxed threshold of –min-score = 0.75 to increase detection sensitivity. The putative viral contigs were annotated using our previously developed ViroProfiler pipeline [ 71 ] for precise viral contig identification and annotation. Briefly, The ViroProfiler pipeline utilizes CheckV (v0.9.0) [ 72 ] to remove host region in provirus contigs, and then use VirSorter2 (v2.2.4) [ 73 ] and VIBRANT (v1.2.1) [ 74 ] to identify viral contigs. Auxiliary metabolic genes (AMGs) were annotated using VIBRANT and DRAM-v (v1.4.4) [ 75 ]. Viral hosts were predicted using iPHoP (v1.2.0) with confidence score ≥ 90 [ 76 ]. Taxonomic classification was performed using a combination of geNomad and VITAP (v1.7.1) [ 77 ]. VITAP taxonomy was preferred due to its more specific annotations, often to genus or species level with high precision, while geNomad taxonomy was used when a contig was not annotated by VITAP. Viruses were classified as temperate if they are identified as provirus by geNomad, VIBRANT, or CheckV (v0.9.0), or classified as temperate phages by BACPHLIP (v0.9.6) [ 78 ]. Viral contigs were then dereplicated to generate a non-redundant catalog of viral sequences. Specifically, pairwise average nucleotide identity (ANI) was calculated among these contigs using the clustering algorithm within CheckV, applying thresholds of ≥ 95% ANI and ≥ 85% AF to define species-level viral operational taxonomic units (vOTUs). The longest contig within each vOTU cluster was selected as its representative sequence. To further investigate relationships among the identified vOTUs, compare them with known viral genomes, and identify potential novel viral clusters (VCs), gene-sharing network analysis was conducted using vConTACT2 (v.0.11.3) [ 79 ]. VCs identified by vConTACT2 represent groups of related viral genomes, offer insights into potential genus-level relationships, and highlight vOTUs that may constitute novel taxa. Abundance of viruses were estimated by mapping clean reads to vOTUs using Bowtie2, and then quantified using CoverM (v0.7.0) ( https://github.com/wwood/CoverM ). For all downstream analyses, including diversity calculations, differential abundance testing, host prediction, and AMG identification, only vOTUs with an estimated completeness ≥ 50% were used. Statistical analyses All statistical analyses were performed in R (v4.1.0). Data manipulation and visualization utilized the tidyverse ecosystem (v1.3.0). Student’s t -test and Wilcoxon rank sum test were employed for group comparisons. The analysis of differentially abundant bacterial species and viral clusters (VCs) was conducted using ANCOM-BC with Benjamini–Hochberg False Discovery Rate (FDR) method. For transcriptome data, differentially expressed genes (DEGs) were identified using DESeq2 with thresholds of adjusted p -value (Benjamini-Hochberg) < 0.05 and |log 2 FC|≥ 0.585. KEGG pathway enrichment of DEGs was performed using ClusterProfiler (v4.0) [ 80 ] and the resulting p -values were adjusted for multiple testing using the FDR correction. For metabolomics data, peak areas were normalized via total sum scaling. Prior to differential abundance analysis, metabolites that were undetected (missing) in more than 50% of samples within both groups were excluded to prevent statistical artifacts and artificial fold-change inflation. Subsequently, differential metabolites between two groups were identified with threshold of |log 2 FC |≥ 1 and p < 0.05. Metabolic pathway enrichment was analyzed in MetaboAnalyst 5.0 ( https://www.metaboanalyst.ca ) with FDR correction. Supplementary Information 40168_2026_2368_MOESM1_ESM.docx (1.2MB, docx) Supplementary Material 1: Supplementary Figure S1: FE traits and SCFAs concentrations between two groups. Supplementary Figure S2: Bacterial composition, network analysis, and functional correlations between two groups. Supplementary Figure S3: Host association and differential abundance of viral clusters (VCs). Supplementary Figure S4: Volcano plot of differentially abundant metabolites in CF vs. ZF. Supplementary Table S1: Growth, feed efficiency and fatness traits in two groups. Supplementary Table S2: Comparative topological properties of the CF and ZF networks. Supplementary Table S3: Degree and eigenvector centrality of the hub nodes in CF and ZF networks. Acknowledgements The authors sincerely thank Prof. Guang Yao from the University of Arizona and Prof. Xujun Liang from Northwest A&F University for their constructive discussions and valuable suggestions. Additionally, we acknowledge the Technical University of Munich (TUM), the Helmholtz Zentrum München (HMGU), and Northwest A&F University (NWAFU) for providing high-performance computing resources. Authors’ contributions All authors contributed intellectually to and agreed to this submission. XW, YSZ, and JLR designed the experiments. SJT, ZZW, XQY, SFL, JW, WZ, SQW, and YSZ conducted the experiments and collected the data. SJT, LY, and JLR analyzed and interpreted the data. SJT, PYL, DH, and YSZ prepared figures. XW interpreted the data and wrote the manuscript, while JLR, FYL, LD and SSH provided substantial feedback. JLR, XW, and PYL contributed to the conceptual design of the study. XW, YSZ, FYL, LD and SSH provided funding support. All authors read and approved the final manuscript. Funding Open Access funding enabled and organized by Projekt DEAL. This work was supported by the National Key Research and Development Program of China (2021YFD1200300) and the National Natural Science Foundation of China (#32341055) awarded to YSZ, the China Agriculture Research System of MOF and MARA (CARS-42–2) awarded to XW, the Fundamental Research Funds for the Central Universities (grant no. 226–2025-00030) awarded to FYL, the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation– Emmy Noether program, Project No. 273124240; and SPP2330, Project No. 464797012), and the European Research Council Starting grant (ERC StG 803077) awarded to LD. Data availability The raw data for multi-omics cohort are deposited in the National Genomics Data Center website ( https://ngdc.cncb.ac.cn ). All sequencing data are accessible with the identifier CRA026543 (amplicon), CRA028005 (metagenome), and CRA027933 (transcriptome). The metabolomics data are accessible with the identifier OMIX011117. The code used for data analysis is available at the GitHub repository https://github.com/rujinlong/duckbiome . Declarations Ethics approval and consent to participate All animal procedures involving Pekin ducks (Anas platyrhynchos domesticus) were approved by the Animal Welfare and Ethics Committee of the Institute of Animal Sciences (IAS), the Chinese Academy of Agricultural Sciences (IAS20160401, CAAS, Beijing, China). Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Xia Wang, Shujie Tian, and Yunsheng Zhang contributed equally to this work. Contributor Information Fuyong Li, Email: [email protected]. Shuisheng Hou, Email: [email protected]. Pengying Li, Email: [email protected]. Jinlong Ru, Email: [email protected]. References 1. Zampiga M, et al. Application of omics technologies for a deeper insight into quali-quantitative production traits in broiler chickens: a review. J Anim Sci Biotechnol. 2018;9:61. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Wrzosek L, et al. Bacteroides thetaiotaomicron and Faecalibacterium prausnitzii influence the production of mucus glycans and the development of goblet cells in the colonic epithelium of a gnotobiotic model rodent. BMC Biol. 2013;11:61. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Han H, et al. From gut microbiota to host appetite: gut microbiota-derived metabolites as key regulators. Microbiome. 2021;9(1):162. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Ikeda T, et al. Short-chain fatty acid receptors and gut microbiota as therapeutic targets in metabolic, immune, and neurological diseases. Pharmacol Ther. 2022;239:108273. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Gao P, et al. Feed-additive probiotics accelerate yet antibiotics delay intestinal microbiota maturation in broiler chicken. Microbiome. 2017;5(1):91. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Yang J, et al. Microbiota-accessible fiber activates short-chain fatty acid and bile acid metabolism to improve intestinal mucus barrier in broiler chickens. Microbiol Spectr. 2024;12(1):e0206523. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Yang C, et al. Metagenomic insights into the relationship between intestinal flora and residual feed intake of meat ducks. Poult Sci. 2024;103(7):103836. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. He Z, et al. Combined effect of microbially derived cecal SCFA and host genetics on feed efficiency in broiler chickens. Microbiome. 2023;11(1):198. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Zhou GC, et al. Multi-omics revealed the mechanism of feed efficiency in sheep by the combined action of the host and rumen microbiota. Anim Nutr. 2024;18:367–79. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Hoyles L, et al. Characterization of virus-like particles associated with the human faecal and caecal microbiota. Res Microbiol. 2014;165(10):803–12. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Lepage P, et al. Dysbiosis in inflammatory bowel disease: a role for bacteriophages? Gut. 2008;57(3):424–5. [ DOI ] [ PubMed ] [ Google Scholar ] 12. Yan M, et al. Interrogating the viral dark matter of the rumen ecosystem with a global virome database. Nat Commun. 2023;14(1):5254. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Campbell DE, et al. Infection with bacteroides phage BV01 alters the host transcriptome and bile acid metabolism in a common human gut microbe. Cell Rep. 2020;32(11):108142. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Thiele Orberg E, et al. Bacteria and bacteriophage consortia are associated with protective intestinal metabolites in patients receiving stem cell transplantation. Nat Cancer. 2024;5(1):187–208. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Mangalea MR, et al. Individuals at risk for rheumatoid arthritis harbor differential intestinal bacteriophage communities with distinct metabolic potential. Cell Host Microbe. 2021;29(5):726-739.e5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Liu X, et al. Rumen DNA virome and its relationship with feed efficiency in dairy cows. Microbiome. 2025;13(1):14. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Peschel S, et al. NetCoMi: network construction and comparison for microbiome data in R. Brief Bioinform. 2021;22(4):bbaa290. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Qin J, et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature. 2012;490(7418):55–60. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Levi Mortera S, et al. Functional and taxonomic traits of the gut microbiota in type 1 diabetes children at the onset: a metaproteomic study. Int J Mol Sci. 2022;23(24):15982. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Wang S, et al. Dose-related effects of early-life intake of sn-2 palmitate, a specific positionally distributed human milk fatty acid, on the composition and metabolism of the intestinal microbiota. J Dairy Sci. 2023;106(12):8272–86. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Hao L, et al. Novel syntrophic bacteria in full-scale anaerobic digesters revealed by genome-centric metatranscriptomics. ISME J. 2020;14(4):906–18. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Hampton HG, Watson BNJ, Fineran PC. The arms race between bacteria and their phage foes. Nature. 2020;577(7790):327–36. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Murphy J, et al. Bacteriophage orphan DNA methyltransferases: insights from their bacterial origin, function, and occurrence. Appl Environ Microbiol. 2013;79(24):7547–55. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. van Kempen M, et al. Fast and accurate protein structure search with Foldseek. Nat Biotechnol. 2024;42(2):243–6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Deramchia K, et al. Contribution of pyruvate phosphate dikinase in the maintenance of the glycosomal ATP/ADP balance in the Trypanosoma brucei procyclic form. J Biol Chem. 2014;289(25):17365–78. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Feng WH, et al. Lytic induction therapy for Epstein-Barr virus-positive B-cell lymphomas. J Virol. 2004;78(4):1893–902. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Westphal EM, et al. Activation of lytic Epstein-Barr virus (EBV) infection by radiation and sodium butyrate in vitro and in vivo: a potential method for treating EBV-positive malignancies. Cancer Res. 2000;60(20):5781–8. [ PubMed ] [ Google Scholar ] 28. Wen C, et al. Joint contributions of the gut microbiota and host genetics to feed efficiency in chickens. Microbiome. 2021;9(1):126. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Liu Y. Fatty acids, inflammation and intestinal health in pigs. J Anim Sci Biotechnol. 2015;6(1):41. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Ji ZQ, et al. Effects of coated sodium butyrate on growth performance, intestinal morphology, and caecal short-chain fatty acids of growing Pekin ducks. Br Poult Sci. 2023;64(4):529–33. [ DOI ] [ PubMed ] [ Google Scholar ] 31. Yang Q, et al. Butyrate in combination with forskolin alleviates necrotic enteritis, increases feed efficiency, and improves carcass composition of broilers. J Anim Sci Biotechnol. 2022;13(1):3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Liu S, et al. Ecological stability of microbial communities in Lake Donghu regulated by keystone taxa. Ecol Indic. 2022;136:108695. [ Google Scholar ] 33. Maslov S, Sneppen K. Specificity and stability in topology of protein networks. Science. 2002;296(5569):910–3. [ DOI ] [ PubMed ] [ Google Scholar ] 34. Wang Y, et al. Effect of host genetics and gut microbiome on fat deposition traits in pigs. Front Microbiol. 2022;13:925200. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Bergamaschi M, et al. Gut microbiome composition differences among breeds impact feed efficiency in swine. Microbiome. 2020;8(1):110. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Yoon JH, et al. Inclusion of Lacticaseibacillus paracasei NSMJ15 in broiler diets induces changes in jejunal immune cell population and cecal microbiota. Anim Biosci. 2024;37(12):2155–66. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Zhong ZP, et al. Viral potential to modulate microbial methane metabolism varies by habitat. Nat Commun. 2024;15(1):1857. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Obeng N, Pratama AA, Elsas JDV. The significance of mutualistic phages for bacterial ecology and evolution. Trends Microbiol. 2016;24(6):440–9. [ DOI ] [ PubMed ] [ Google Scholar ] 39. Luo XQ, et al. Viral community-wide auxiliary metabolic genes differ by lifestyles, habitats, and hosts. Microbiome. 2022;10(1):190. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Feng T, et al. Metabolite profile of sheep serum with high or low average daily gain. Front Vet Sci. 2021;8:662536. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Wang B, et al. Impact of feeding regimens on the composition of gut microbiota and metabolite profiles of plasma and feces from Mongolian sheep. J Microbiol. 2020;58(6):472–82. [ DOI ] [ PubMed ] [ Google Scholar ] 42. Li S, et al. Longitudinal changes of serum metabolomic profile after laparoscopic sleeve gastrectomy in obesity. Endocr Connect. 2024;13(11):e240292. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Han Z, et al. Fecal microbiota transplantation accelerates restoration of florfenicol-disturbed intestinal microbiota in a fish model. Commun Biol. 2024;7(1):1006. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Li Z, et al. Traditional Tibetan medicine twenty-five wei’er tea pills ameliorate rheumatoid arthritis based on chemical crosstalk between gut microbiota and the host. Front Pharmacol. 2022;13:828920. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Zheng X, et al. Integrated metagenomic and metabolomics profiling reveals key gut microbiota and metabolites associated with weaning stress in piglets. Genes. 2024. 10.3390/genes15080970. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Huang Z, et al. Gut microbiota and blood metabolites: unveiling their roles in hippocampal volume changes through Mendelian randomization and mediation analysis. Metab Brain Dis. 2025;40(4):178. [ DOI ] [ PubMed ] [ Google Scholar ] 47. Yan J, et al. Fecal microbiota transplantation significantly improved respiratory failure of amyotrophic lateral sclerosis. Gut Microbes. 2024;16(1):2353396. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Zhang Y, et al. Genetic parameters for residual feed intake in a random population of Pekin duck. Asian-Australas J Anim Sci. 2017;30(2):167–70. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Chen S, et al. Fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Dobin A, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 52. Callahan BJ, et al. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581–3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Bolyen E, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019;37(8):852–7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 54. McMurdie PJ, Holmes S. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8(4):e61217. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 55. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 56. Nurk S, et al. metaSPAdes: a new versatile metagenomic assembler. Genome Res. 2017;27(5):824–34. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 57. Kang DD, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ. 2019;7:e7359. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 58. Wu YW, Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics. 2016;32(4):605–7. [ DOI ] [ PubMed ] [ Google Scholar ] 59. Alneberg J, et al. Binning metagenomic contigs by coverage and composition. Nat Methods. 2014;11(11):1144–6. [ DOI ] [ PubMed ] [ Google Scholar ] 60. Sieber CMK, et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nat Microbiol. 2018;3(7):836–43. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 61. Chklovski A, et al. CheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning. Nat Methods. 2023;20(8):1203–12. [ DOI ] [ PubMed ] [ Google Scholar ] 62. Chaumeil PA, et al. GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database. Bioinformatics. 2019;36(6):1925–7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 63. Beghini F, et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. Elife. 2021;10:e65088. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 64. Lin H, Peddada SD. Analysis of compositions of microbiomes with bias correction. Nat Commun. 2020;11(1):3514. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Hyatt D, et al. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics. 2010;11:119. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 66. Cantalapiedra CP, et al. eggNOG-mapper v2: functional annotation, orthology assignments, and domain prediction at the metagenomic scale. Mol Biol Evol. 2021;38(12):5825–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 67. Li W, Godzik A. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics. 2006;22(13):1658–9. [ DOI ] [ PubMed ] [ Google Scholar ] 68. Patro R, et al. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017;14(4):417–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 69. Mallick H, et al. Multivariable association discovery in population-scale meta-omics studies. PLoS Comput Biol. 2021;17(11):e1009442. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 70. Camargo AP, et al. Identification of mobile genetic elements with geNomad. Nat Biotechnol. 2024;42(8):1303–12. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 71. Ru J, et al. Viroprofiler: a containerized bioinformatics pipeline for viral metagenomic data analysis. Gut Microbes. 2023;15(1):2192522. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 72. Nayfach S, et al. Checkv assesses the quality and completeness of metagenome-assembled viral genomes. Nat Biotechnol. 2021;39(5):578–85. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 73. Guo J, et al. VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome. 2021;9(1):37. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 74. Kieft K, Zhou Z, Anantharaman K. VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences. Microbiome. 2020;8(1):90. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. Shaffer M, et al. DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res. 2020;48(16):8883–900. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 76. Roux S, et al. IpHoP: an integrated machine learning framework to maximize host prediction for metagenome-derived viruses of archaea and bacteria. PLoS Biol. 2023;21(4):e3002083. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 77. Zheng K, et al. VITAP: a high precision tool for DNA and RNA viral classification based on meta-omic data. Nat Commun. 2025;16(1):2226. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 78. Hockenberry AJ, Wilke CO. BACPHLIP: predicting bacteriophage lifestyle from conserved protein domains. PeerJ. 2021;9:e11396. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 79. Bin Jang H, et al. Taxonomic assignment of uncultivated prokaryotic virus genomes is enabled by gene-sharing networks. Nat Biotechnol. 2019;37(6):632–9. [ DOI ] [ PubMed ] [ Google Scholar ] 80. Wu T, et al. ClusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021;2(3):100141. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials 40168_2026_2368_MOESM1_ESM.docx (1.2MB, docx) Supplementary Material 1: Supplementary Figure S1: FE traits and SCFAs concentrations between two groups. Supplementary Figure S2: Bacterial composition, network analysis, and functional correlations between two groups. Supplementary Figure S3: Host association and differential abundance of viral clusters (VCs). Supplementary Figure S4: Volcano plot of differentially abundant metabolites in CF vs. ZF. Supplementary Table S1: Growth, feed efficiency and fatness traits in two groups. Supplementary Table S2: Comparative topological properties of the CF and ZF networks. Supplementary Table S3: Degree and eigenvector centrality of the hub nodes in CF and ZF networks. Data Availability Statement The raw data for multi-omics cohort are deposited in the National Genomics Data Center website ( https://ngdc.cncb.ac.cn ). All sequencing data are accessible with the identifier CRA026543 (amplicon), CRA028005 (metagenome), and CRA027933 (transcriptome). The metabolomics data are accessible with the identifier OMIX011117. The code used for data analysis is available at the GitHub repository https://github.com/rujinlong/duckbiome . Articles from Microbiome are provided here courtesy of BMC ACTIONS View on publisher site PDF (6.0 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top