Metagenomics of Virus Diversities in Solid-State Brewing Process of Traditional Chinese Vinegar

Traditional Chinese vinegar offers an exceptional flavor and rich nutrients due to its unique solid-state fermentation process, which is a multiple microbial fermentation system including various bacteria, fungi and viruses. However, few studies on the virus diversities in traditional Chinese vinegar have been reported. In this paper, using Zhenjiang aromatic vinegar as a model system, we systemically explored the viral communities in the solid-state brewing process of traditional Chinese vinegar using bacterial and viral metagenomes. Results showed that the viral diversity in vinegar Pei was extensive and the virus communities varied along with the fermentation process. In addition, there existed some interactions between viral and bacterial communities. Moreover, abundant antibiotic resistance genes were found in viromes, indicating that viruses might protect fermentation bacteria strains from the stress of antibiotics in the fermentation environment. Remarkably, we identified abundant auxiliary carbohydrate metabolic genes (including alcohol oxidases, the key enzymes for acetic acid synthesis) from viromes, implying that viruses might participate in the acetic acid synthesis progress of the host through auxiliary metabolic genes. Taken together, our results indicated the potential roles of viruses in the vinegar brewing process and provided a new perspective for studying the fermentation mechanisms of traditional Chinese vinegar.


Introduction
Traditional Chinese vinegar has played an essential role in Chinese life through the ages for its unique flavor, high nutrition (such as polyphenols, organic acids, melanoidins and tetramethylpyrazine) and many important biological functions, including antioxidative activity, liver protection, blood pressure and glucose control [1], lipid metabolism regulation and anti-tumor [2]. The main characteristic of traditional Chinese vinegar is the unique multispecies solid-state fermentation process, which is conducted in an open fermentation environment containing many microorganisms, including viruses, bacteria and fungi [3]. Presently, there were many studies reported on the bacteria and fungi in traditional Chinese vinegar fermentation, but few studies focused on viruses.
Viruses are hypothesized to be a major driver of their hosts' evolution [4]. In general, viruses such as bacteriophages are considered harmful to fermented food production by decreasing the fermentative capacity of fermentative strains, occasionally resulting in complete fermentation failure. However, the diversity of viruses varied in diverse fermented foods due to the different environments. It has been observed that some viruses can positively influence the fermentation process of fermented foods [5]. Reports showed that, in some fermented food such as cocoa beans and milk cheese, viruses can regulate

Virus Purification
The viruses in vinegar Pei were purified according to the method of Dugat-Bony [19] with slight modification. Briefly, 5 g vinegar Pei samples were put into a sterile bag. Then, 50 mL cold sterile 2% (w/v) trisodium citrate was added and mixed for 3 min using a flapping sterile homogenizer (Shanghai Lichen Bangxi Instrument Technology Co., Ltd., Shanghai, China). After centrifuging at 500× g, 4 °C for 10 min, big aggregates in the solution were discarded. The supernatant was subsequently centrifuged at 5000× g, 4 °C for 10 min to remove the microbial cells and the free viral particles were in the supernatant. Then, the precooled SM buffer (200 mM NaCl, 10 mM MgSO4, 50 mM Tris pH 7.5) supernatant was used to dilute the supernatant in a ratio of SM buffer: supernatant = 5:1 (v/v). Next, the solution was filtrated using 0.22 μm polyethersulfone membranes. RNase A and DNase Ⅰ were added into the filtrate with the final concentration of 1 μg/mL and left at 37 °C for 30 min to degrade genomic DNA and RNA of bacterial cells. Then, cold sterile PEG 8000 (Sigma) aqueous solution was added to the solution to make a final concentration 10% (w/v) and kept overnight at 4 °C to precipitate viral particles. The solution was subsequently centrifuged at 12,000× g, 4 °C for 1 h and the supernatant was discarded. Finally, virus particles in bottom sediment were resuspended with 2 mL of cold SM buffer and then stocked at 4 °C until ready to use.

Viral and Bacterial DNA Extraction and Virome Sequencing
Genomic DNA of purified viruses in vinegar Pei was extracted using a Magnetic Virus DNA/RNA Kit (Tiangen Biotech (Beijing) Co., Ltd., Beijing, China) following the manufacturer's protocol. Genomic DNA of bacteria in vinegar Pei was extracted using a TIANamp Bacteria DNA Kit (Tiangen Biotech (Beijing) Co., Ltd., Beijing, China) following the manufacturer's protocol. DNA libraries were constructed with an insert size of 300 bp and sequenced using an Illumina Hiseq PE150 sequencing platform (Illumina Inc., San Diego, CA, USA) at the Shanghai Rongxiang Biotechnology Co., Ltd. (Shanghai, China).

Virus Purification
The viruses in vinegar Pei were purified according to the method of Dugat-Bony [19] with slight modification. Briefly, 5 g vinegar Pei samples were put into a sterile bag. Then, 50 mL cold sterile 2% (w/v) trisodium citrate was added and mixed for 3 min using a flapping sterile homogenizer (Shanghai Lichen Bangxi Instrument Technology Co., Ltd., Shanghai, China). After centrifuging at 500× g, 4 • C for 10 min, big aggregates in the solution were discarded. The supernatant was subsequently centrifuged at 5000× g, 4 • C for 10 min to remove the microbial cells and the free viral particles were in the supernatant. Then, the precooled SM buffer (200 mM NaCl, 10 mM MgSO 4 , 50 mM Tris pH 7.5) supernatant was used to dilute the supernatant in a ratio of SM buffer: supernatant = 5:1 (v/v). Next, the solution was filtrated using 0.22 µm polyethersulfone membranes. RNase A and DNase I were added into the filtrate with the final concentration of 1 µg/mL and left at 37 • C for 30 min to degrade genomic DNA and RNA of bacterial cells. Then, cold sterile PEG 8000 (Sigma) aqueous solution was added to the solution to make a final concentration 10% (w/v) and kept overnight at 4 • C to precipitate viral particles. The solution was subsequently centrifuged at 12,000× g, 4 • C for 1 h and the supernatant was discarded. Finally, virus particles in bottom sediment were resuspended with 2 mL of cold SM buffer and then stocked at 4 • C until ready to use.

Viral and Bacterial DNA Extraction and Virome Sequencing
Genomic DNA of purified viruses in vinegar Pei was extracted using a Magnetic Virus DNA/RNA Kit (Tiangen Biotech (Beijing) Co., Ltd., Beijing, China) following the manufacturer's protocol. Genomic DNA of bacteria in vinegar Pei was extracted using a TIANamp Bacteria DNA Kit (Tiangen Biotech (Beijing) Co., Ltd., Beijing, China) following the manufacturer's protocol. DNA libraries were constructed with an insert size of 300 bp and sequenced using an Illumina Hiseq PE150 sequencing platform (Illumina Inc., San Diego, CA, USA) at the Shanghai Rongxiang Biotechnology Co., Ltd. (Shanghai, China). Raw data produced were filtered to remove reads containing "N" bases (N parameter setting 10), reads with adaptors and low-quality reads with a quality score of <20 using Cutadapt software version 1.18 [20], retaining high-quality clean data for subsequent analysis. All virome reads were assembled into contigs using Megahit software version 1.1.3 [21]. Moreover, open reading frames (ORFs) were predicted using MetaGeneMark [22].

Identification of Auxiliary Carbohydrate Metabolic Genes
According to the method of Jin [23], in order to obtain the clusters of orthologous groups of proteins (COG) corresponding to vinegar Pei virus genes, COG functional annotation of vinegar Pei viromes was performed using protein-protein BLAST (BLASTp) software [24]. Predicted ORFs of vinegar Pei viromes were compared with eggNOG database (http://eggnog5.embl.de/ (accessed on 30 November 2021)) with an e-value threshold of 1 × 10 −5 . Furthermore, ORFs related to carbohydrate metabolism of vinegar Pei viromes were obtained from ORFs belonging to the COG function class of carbohydrate transport and metabolism using CD-Hit software [25] with thresholds of 95% identity plus 90% coverage. Subsequently, carbohydrate-active enzymes (CAZymes) from these viral ORFs were identified by the hmmscan program from HMMER v.3.1 [26] compared with CAZymes database using e-value ≤ 1 × 10 −5 as a cut-off. In addition, in order to receive the best annotation of each ORF, ORFs associated with carbohydrate metabolism and CAZymes were compared with the National Center for Biotechnology Information (NCBI) NR and Pfam database.

Antibiotic Resistance Gene Search
The CDS of bacterial and viral metagenomes ORFs were used as queries to search for ARGs in vinegar Pei virome using Diamond software [27] against the comprehensive antibiotic resistance database (CARD) with the comparison parameter was set to 'strict'.

Investigation of Bacterial and Viral Metagenomes of Vinegar Pei
We defined the reads with a quality score of ≥20, no adaptors and no ambiguous 'N' bases as high-quality reads. In bacterial metagenomics, an average of 11.16 gigabase pairs (Gbp) with GC content of 55.58% and 76.25 million high-quality reads were obtained based on Q20% > 90. A total of 1,136,418 contigs were assembled from 12 DNA samples extracted from vinegar Pei bacteria (Table 1) and gene sequence length averaged 1044 bp (N50: 1193 bp). In viral metagenomics, an average of 1.6 Gbp with GC content of 54.43% and 11.05 million high-quality reads were acquired based on Q20% > 90. Gene sequence length averaged 1080 bp (N50: 1264 bp). Moreover, a total of 150,703 contigs were assembled from 12 DNA samples extracted from vinegar Pei virus (Table 1). Moreover, we made an ORFs prediction analysis using MetaGeneMark software. One hundred and thirty-three thousand nine hundred and fifty-nine and twenty thousand five hundred and fifteen ORFs were obtained in metagenomics and viral metagenomics of vinegar Pei samples, respectively ( Table 1). The above results showed that the throughput and sequencing quality of bacterial and viral metagenomes sequencing data of vinegar Pei were high enough for the following analyses.

Taxonomic Diversities of Vinegar Pei Viral and Bacterial Communities
Bacterial taxonomic affiliations of vinegar Pei were determined by comparing the predicted bacterial genomic ORFs with bacterial sequences from the NCBI RefSeq Bacteria database. Results showed that a total of 532 bacterial families were identified in vinegar Pei during acetic acid fermentation (Supplementary File S1). Acetomonas accounted for the largest fraction, and the following in order were Xanthomonaceae, Sphingomonas, Komagataeibacter and Oligotrophiaceae. In addition, 1668 bacterial genera were identified in vinegar Pei (Figure 2a). Moreover, we made an ORFs prediction analysis using MetaGeneMark software. One hundred and thirty-three thousand nine hundred and fifty-nine and twenty thousand five hundred and fifteen ORFs were obtained in metagenomics and viral metagenomics of vinegar Pei samples, respectively ( Table 1). The above results showed that the throughput and sequencing quality of bacterial and viral metagenomes sequencing data of vinegar Pei were high enough for the following analyses.

Taxonomic Diversities of Vinegar Pei Viral and Bacterial Communities
Bacterial taxonomic affiliations of vinegar Pei were determined by comparing the predicted bacterial genomic ORFs with bacterial sequences from the NCBI RefSeq Bacteria database. Results showed that a total of 532 bacterial families were identified in vinegar Pei during acetic acid fermentation (Supplementary File S1). Acetomonas accounted for the largest fraction, and the following in order were Xanthomonaceae, Sphingomonas, Komagataeibacter and Oligotrophiaceae. In addition, 1668 bacterial genera were identified in vinegar Pei (Figure 2a).  Furthermore, by comparing the virome ORFs to those in the NCBI RefSeq Virus database, the viral taxonomic composition of vinegar Pei was obtained. Results showed that only a tiny fraction of the virome ORFs was similar to some sequences in the NCBI RefSeq Virus database, while most of the vinegar Pei viruses were unknown. Finally, a total of 40 viral families were identified in vinegar Pei viruses (Supplementary File S2). Myoviridae accounted for the largest fraction in vinegar Pei and Siphoviridae, Caudovirales were, respectively, in the second and third position. In addition, 258 viral genera were identified in vinegar Pei (Figure 2b).

Dynamic Changes of Viral Communities during the Fermentation and the Interaction with Bacterial Community in Vinegar Pei
The virus communities in vinegar Pei varied with the fermentation process ( Figure 3b). Compared with the fermentation at 0d, the virus community structure changed significantly after the beginning of fermentation. On day 0 of fermentation, Badnavirus and Errantivirus were the main two virus genus, which might be from the fermentation raw materials (rice wine, rice husk and bran), while, on the 8th day of fermentation, the top five virus genus were Myoviridae, Siphoviridae, Caudovirales, Bcep78 and Pbunavirus. In addition, the community structure of viruses in vinegar Pei also changed along with fermentation. Compared with the 8th day of fermentation, the abundance of Myoviridae and Siphoviridae increased gradually, while Badnavirus and Errantivirus decreased and were not detected on the 18th day of fermentation. Bcep78, Pbunavirus and Phietavirus increased firstly, then reduced. Podoviridae decreased firstly and then increased. Caudovirales remained unchanged at first and gradually increased on the 12th day of fermentation. However, the change in P1 virus was insignificant.
The virus communities in vinegar Pei varied with the fermentation process ( Figure  3b). Compared with the fermentation at 0d, the virus community structure changed significantly after the beginning of fermentation. On day 0 of fermentation, Badnavirus and Errantivirus were the main two virus genus, which might be from the fermentation raw materials (rice wine, rice husk and bran), while, on the 8th day of fermentation, the top five virus genus were Myoviridae, Siphoviridae, Caudovirales, Bcep78 and Pbunavirus. In addition, the community structure of viruses in vinegar Pei also changed along with fermentation. Compared with the 8th day of fermentation, the abundance of Myoviridae and Siphoviridae increased gradually, while Badnavirus and Errantivirus decreased and were not detected on the 18th day of fermentation. Bcep78, Pbunavirus and Phietavirus increased firstly, then reduced. Podoviridae decreased firstly and then increased. Caudovirales remained unchanged at first and gradually increased on the 12th day of fermentation. However, the change in P1 virus was insignificant. To better understand the interactions between virus and bacteria in vinegar Pei, the changes of bacterial community structure during acetic acid fermentation were also analyzed using metagenomics. It was found that from the 8th day of fermentation, Acetomonas was the dominant bacteria. According to the abundance value, the top five genera were Acetomonas, Xanthomonas, Sphingomonas, Komagataeibacter and Stenotrophomonas. Among them, Acetomonas was the main dominant bacteria. By the 18th day of fermentation, the abundance of Acetomonas reached the maximum (78.4%). The changes in virus community and microbial community were preliminarily analyzed. The results showed that Myoviridae, Siphoviridae and Caudovirales were consistent with the growth trend of Acetomonas, while Badnavirus and Errantivirus showed the opposite trend to that of Acetomonas, indicating that Myoviridae, Siphoviridae, Caudovirales, Badnavirus and Errantivirus might play an essential role in the changes of Acetomonas community and metabolism. Consistently, Koki Omata [28] obtained some temperate phages from acetic acid bacteria induced by mitomycin C, and transmission electron microscopy and genomic analysis revealed that all of these belong to the myoviridae-type phage, indicating that acetic acid bacteria was the host of the myoviridae-type phage and the latter might affect the metabolism of acetic acid bacteria. To better understand the interactions between virus and bacteria in vinegar Pei, the changes of bacterial community structure during acetic acid fermentation were also analyzed using metagenomics. It was found that from the 8th day of fermentation, Acetomonas was the dominant bacteria. According to the abundance value, the top five genera were Acetomonas, Xanthomonas, Sphingomonas, Komagataeibacter and Stenotrophomonas. Among them, Acetomonas was the main dominant bacteria. By the 18th day of fermentation, the abundance of Acetomonas reached the maximum (78.4%). The changes in virus community and microbial community were preliminarily analyzed. The results showed that Myoviridae, Siphoviridae and Caudovirales were consistent with the growth trend of Acetomonas, while Badnavirus and Errantivirus showed the opposite trend to that of Acetomonas, indicating that Myoviridae, Siphoviridae, Caudovirales, Badnavirus and Errantivirus might play an essential role in the changes of Acetomonas community and metabolism. Consistently, Koki Omata [28] obtained some temperate phages from acetic acid bacteria induced by mitomycin C, and transmission electron microscopy and genomic analysis revealed that all of these belong to the myoviridae-type phage, indicating that acetic acid bacteria was the host of the myoviridae-type phage and the latter might affect the metabolism of acetic acid bacteria.

Abundant Auxiliary Carbohydrate Metabolic Genes in Vinegar Pei Viruses
Viruses could regulate host metabolism by encoding auxiliary metabolic genes (AMGs), a series of homologous genes related to host metabolism [10,12]. Moreover, AMGs encoded by viruses have different characteristics and advantages from host homologous genes [29]. The raw materials of traditional Chinese vinegar are mainly carbohydrate-rich cereals such as rice and sorghum [2], and there are many carbohydrate catabolism and anabolism reactions that happen in traditional Chinese vinegar brewing, which need many carbohydrate-metabolism enzymes. In order to investigate whether vinegar Pei viruses contain auxiliary carbohydrate metabolic genes or not, vinegar Pei virome sequences were compared with the eggNOG database to obtain the clusters of orthologous groups of pro-  Figure 4a showed that most viral ORFs were not annotated, revealing that there were a large number of uncharacterized viral genes in vinegar Pei. Although annotated viral ORFs were grouped into all COG functional categories, most ORFs were associated to conventional viral functions such as 'replication, recombination and repair', 'amino acid transport and metabolism', 'translation, ribosomal structure and biogenesis', 'energy production and conversion', 'nucleotide transport and metabolism'. Notably, besides the above-mentioned, COG functional categories for 'carbohydrate transport and metabolism' were significantly over-represented in vinegar Pei virome (Figure 4a).  In the carbohydratae-active enzyme (CAZymes) database, carbohydrate-active enzymes from different species can be separated as glycoside hydrolases (GHs) family, glycosyltransferases (GTs) family, polysaccharide lyases (PLs) family, carbohydrate esterases (CEs) family, carbohydrate-binding modules (CBMs) family and auxiliary activities (AAs) family. Using hmmscan program from HMMER v.3.1 software, the vinegar Pei viral ORFs were compared with the CAZymes database. Results showed that, in vinegar Pei viruses, CEs, GHs and GTs family were the top three auxiliary carbohydrate-metabolism genes, and the highest gene abundance was obtained on the 8th day of fermentation (Figure 4b). Consistently, the total acid contents of vinegar Pei increased rapidly from the 8th day of In the carbohydratae-active enzyme (CAZymes) database, carbohydrate-active enzymes from different species can be separated as glycoside hydrolases (GHs) family, glycosyltransferases (GTs) family, polysaccharide lyases (PLs) family, carbohydrate esterases (CEs) family, carbohydrate-binding modules (CBMs) family and auxiliary activities (AAs) family. Using hmmscan program from HMMER v.3.1 software, the vinegar Pei viral ORFs were compared with the CAZymes database. Results showed that, in vinegar Pei viruses, CEs, GHs and GTs family were the top three auxiliary carbohydrate-metabolism genes, and the highest gene abundance was obtained on the 8th day of fermentation (Figure 4b). Consistently, the total acid contents of vinegar Pei increased rapidly from the 8th day of fermentation (Supplementary Figure S1), indicating that fermentation metabolism became vigorous from the 8th day of fermentation, requiring more carbohydrate-metabolism enzymes to produce more acetic acid.
Moreover, a total of 127 ORFs were further identified as CAZymes, while 43 auxiliary carbohydrate-metabolism genes were identified in vinegar Pei viruses, which belong to AAs, CEs, GHs, GTs and PLs family, including the common alcohol oxidases, acetylesterase, cellulase, peptidoglycan lytic transglycosylases, mannanase, chitinase, glucosyltransferase, starch phosphorylase, glucan synthase and pectate lyase ( Table 2). Among them, alcohol oxidases are key enzymes in acetic acid synthesis [30]. Cellulose-rich materials such as wheat bran and rice hulls are raw materials of Zhenjiang aromatic vinegar in the acetic acid fermentation process. Meanwhile, Takahashi reported that cellulase might increase the output of vinegar [31]. Presently, the results implied that viruses in vinegar Pei might play essential roles in acetic acid metabolisms through these auxiliary carbohydrate metabolic genes in the acetic acid fermentation process of traditional Chinese vinegar. Consistently, auxiliary carbohydrate metabolic genes were also found in viruses from other ecosystems [23]. The latest research showed that virus-encoded genes related to carbon metabolism (talC, cp12) were found in a cyanophage S-SZBM1 [13]. Table 2. Annotated auxiliary CAZymes from vinegar Pei viruses.

Abundant Antibiotic Resistance Genes in Vinegar Pei Viruses
Viruses were considered as crucial reservoirs of antibiotic resistance genes in the environment [32]. A recent study on the river ecosystem reported that bacteriophages played an important role in antibiotic resistance genes dissemination [33]. Presently, whether ARGs existed in vinegar Pei viromes was also investigated. Results showed that ARGs were obtained in vinegar Pei viromes, which mainly included genes for aminoglycoside, novobiocin, multidrug transport proteins, polymyxin, sulfonamide, fluoroquinolone, elfamycin, mupirocin and quinolone resistance proteins (Table 3). Table 3. Number of virome reads that were predicted to confer antibiotic resistance.
Aminoglycoside resistance proteins function mainly by inactivating aminoglycosides via enzymatic modification of the antibiotic chemical structure [34] or preventing aminoglycosides from binding to the ribosome [35]. Novobiocin resistance proteins found in vinegar Pei viruses were alanyl-tRNA synthetase (alaS). Multidrug transport proteins were antibiotic efflux complex, including ABC-type multidrug efflux pump components [36][37][38]. Polymyxin resistance proteins including ArnA, ArnC and ArnT are required for the synthesis and transfer of 4-amino-4-deoxy-L-arabinose (Ara4N) to Lipid A. Previous reports showed that one of the mechanisms of polymyxins' resistance was the regulatory network controlling chemical modifications of lipid A moiety anchored on lipopolysaccharide, reducing the negative charge of lipid A and its affinity for polymyxins [39]. Sulfonamide resistance proteins were sulfonamide resistant dihydropteroate synthase, which was encoded by two known genes sulI and sulII [40]. Meanwhile, genes at low frequencies encoding resistance proteins for fluoroquinolone, elfamycin, mupirocin and quinolone were also obtained in the vinegar Pei viromes (Table 3).
Overall, the proportions of ARGs in vinegar Pei viral communities ranged from approximately 0 to 0.0062% (Table 3). The proportions of ARGs in the vinegar Pei viromes were comparable with ARG levels of viruses detected in other studies. Bioinformatics analysis of public data downloaded from NCBI RefSeq Protein Database showed that the mean proportion of predicted ARGs found in prophages from natural environments was 0-0.0028% [32]. Moreover, a more variable but still low proportion (0.07-0.12%) of virome reads were annotated as ARGs in the river ecosystem viromes [36].
By comparing the predicted ORFs of the viral contigs against the Comprehensive Antibiotic Resistance Database (CARD), ARGs in vinegar Pei viromes were obtained. Results showed that a total of 29 ARGs were found in vinegar Pei viromes (Table 4). Some ARGs conferred antibiotic resistance by forming an antibiotic efflux complex. lrfA, which is a well-characterized M. smegmatis efflux pump [41] and is involved in the active efflux of quinolones, was found in five vinegar Pei viral contigs. Multidrug transport proteins macA were found in six vinegar Pei viral contigs and macB were found in two vinegar Pei viral contigs. Moreover, APH (3')-IIc, encoded by the ORFs 12d-2-k141-17472 gene 21,035 and 12d-2-k141-18203 gene 22034, is a chromosomal-encoded aminoglycoside phosphotransferase [42] and could inactivate aminoglycoside. Some ARGs, such as EF-Tu, parC and gyrB, conferred antibiotic resistance through antibiotic resistant gene variants or mutants (Table 4). Additionally, alanyl-tRNA synthetase (alaS), an aminocoumarin resistance gene, were found in eight vinegar Pei viral contigs. IleS, a mupirocin resistance gene, was found in three vinegar Pei viral contigs. Moreover, datasets for bacterial community and metagenomes in vinegar Pei were also obtained in parallel at identical sampling stations. Results showed that the ARGs genes above-mentioned were also detected in bacterial metagenomes (Supplementary File S3), revealing that fermentation bacteria strains might be protected by virus-encoded ARGs genes from the stress of antibiotics in the fermentation environment.

Conclusions
Studies on the microenvironment of the solid-state fermentation process are very important for clarifying the fermentation mechanisms of traditional Chinese vinegar. In the present study, using Zhenjiang aromatic vinegar as a model system, we systemically explored the viral communities in the solid-state brewing process of traditional Chinese vinegar for the first time. The results revealed extensive viral diversity in vinegar Pei, which varied with the fermentation process. Meanwhile, there existed some interactions between viral and bacterial communities in vinegar Pei. Moreover, we identified abundant antibiotic resistance genes and auxiliary carbohydrate metabolic genes (including alcohol oxidases, the key enzymes in acetic acid synthesis) from vinegar Pei viromes, indicating the potentially important roles of viruses in traditional Chinese vinegar brewing. In a word, our results provided a new perspective for studying the fermentation mechanisms of traditional Chinese vinegar. However, there is still so much work to be undertaken. Primarily, viruses encoding AMGs in traditional Chinese vinegar brewing would need to be isolated and the functions of these AMGs in the acetic acid fermentation process would need to be verified in the future.