ODFM, an omics data resource from microorganisms associated with fermented foods

Advances in next-generation sequencing technology have led to the rapid expansion of microbial genome sequence data. Easy access, as well as convenient analytical tools, have enabled the exploration of microbial communities in various environmental samples. However, efficient resource usage is becoming increasingly difficult because of the rapid accumulation of sequencing data. Environmental microbiomes in fermented foods, the mammalian gut, and soils comprise not only bacteria, but also archaea, eukaryotic microorganisms, and viruses. These microbial entities are all essential in determining the microbial signature and thus, the inherent characteristics of a given ecosystem. In this context, a comprehensive database covering all genomes of a microbiome within a specific ecosystem would aid in improving our understanding of the complex interactions among the microbial populations.

Fermented foods are an integral part of the global human diet. Microbial entities in fermented foods include bacteria, archaea, yeasts, and viruses. Microbial activities, as well as the type of raw materials, ultimately determine the nutritional and organoleptic properties, quality, and safety of the fermentation product. Given that consumers and manufacturers alike are interested in tasty, high-quality foods as well as the reliability of geographic origins, providing standardised microbial profiles and/or genome information for key microorganisms during the fermentation process is important for ensuring the high quality of fermentation products.

Kimchi is a traditional Korean food prepared by fermentation of vegetables, such as kimchi cabbage, along with various added ingredients and seasonings. The global annual consumption of kimchi is 1,500,000 tons. Like other fermented foods, kimchi shows the presence of a distinct microbial community. Taxonomic studies using culture-dependent and -independent approaches have revealed that lactic acid bacteria (LAB), including Leuconostoc, Lactobacillus, and Weissella, are mainly responsible for kimchi fermentation.

System design and architecture: ODFM is a web-based application compliant with HTML5 and designed using a REST architecture. It is hosted on four servers (web, web application, database, storage) to support stable cloud-based service. The client uses AngularJS (v1.7); the server uses Java Spring framework with integrated Python and FastQC; the database is MySQL. The user interface provides modules for Registration, Data Search, Tools, Our Projects, Statistics, and Q&A.

Data registration and curation: Submitted genome data are stored on the storage server and metadata on the database server. A system administrator verifies, approves, and releases registered data. Automated back-end processing validates file formats, converts data, and generates quality control outputs. JBrowse and GView are integrated for genome visualization. Users can access sequence files, annotations, and QC results via the web interface.

Search and browsing: The platform supports simple (exact) and lexical (partial) keyword searches and allows querying by taxonomy (bacteria, archaea, eukaryotic microorganisms, viruses), study, and sample (e.g., kimchi, fermented seafood, solar salt, soybean paste, vinegar, beer, cheese, sake, yogurt). Results are shown in tables with expandable details on submitters, isolation sources, sequencing, and annotation. Genome browsing provides raw data info, FastQC reports, sequence/annotation details from GenBank-format files, and linear (JBrowse) and circular (GView) genome views derived from GFF files.

Analytical tools: The BLAST suite (blastn, blastp, blastx, tblastn, tblastx) enables local alignments against ODFM DNA/protein databases with standard statistics (bit score, E-value) and downloadable annotation details. For genome relatedness, the dRep tool is integrated to compute Average Nucleotide Identity (ANI). Users create comparative datasets from registered genomes or by uploading FASTA files; query genomes are fragmented (~1,020 bp), MinHash distance (Mash) guides comparisons, and ANI results are returned as downloadable tables and images.

Archive expansion and integrations: ODFM accommodates deposition of fermentative omics data upon request. It integrates outputs from external resources such as KEGG and COGs for functional annotation. A curated list of relevant studies is provided under the ‘Our projects’ section.

• Comprehensive content: 131 bacterial genomes (62 complete, 69 draft) spanning 38 genera and 96 (sub)species; 38 archaeal genomes (7 complete, 31 draft) across 19 genera and 36 species; 28 eukaryotic microbial genomes (14 complete, 14 draft) across 9 species, including spoilage yeasts (Candida, Hanseniaspora, Kazachstania, Pichia, Yarrowia) and species from beer, cheese, and sake ecosystems.
• Community datasets: 70 metagenomes (10 total metagenomes and 60 viral metagenomes from kimchi), 113 bacterial metataxonomes (kimchi, fermented seafood, soybean paste), 9 (meta)transcriptomes (5 kimchi metatranscriptomes and 4 archaeal transcriptomes), and 7 metabolomes (kimchi, fermented seafood). Viral metagenomes available under PRJEB23957.
• Functionality: Integrated BLAST for nucleotide/protein similarity search and dRep-based ANI for genome relatedness and clustering, with user-friendly upload/selection workflows and downloadable outputs. Genome visualization via JBrowse (linear) and GView (circular) supports exploration of annotations and features.
• System environment: CentOS 6.5, JDK 1.8, Apache 2.2.15, Tomcat 7.0, MySQL 5.7; browsing via JBrowse and GView.
• Practical impact: Provides a curated, fermentation-focused resource enabling rapid identification, comparative genomics, and evaluation of candidate starter strains with desirable metabolic traits while supporting safety assessment through genome-based screening.

ODFM consolidates multi-omic datasets for microbial communities associated with fermented foods, delivering both archival and analytical capabilities. By focusing on bacteria, archaea, eukaryotic microorganisms, and viruses within specific food ecosystems (e.g., kimchi, fermented seafood, dairy, and alcoholic fermentations), it addresses the need for ecosystem-specific, comprehensive genomic references. The integrated BLAST and ANI tools streamline initial identification and genome relatedness assessments, facilitating the selection and evaluation of candidate starter cultures. This supports linking microbial taxa to functional attributes relevant to fermentation quality and flavor (e.g., LAB producing mannitol, gamma-aminobutyric acid, lactate, and aroma compounds). While omics resources and tools significantly aid candidate selection, fermentation outcomes remain context-dependent, influenced by environmental factors and microbial interactions. ODFM’s accessible interface and downloadable data enhance reproducibility, sharing, and downstream analyses, and the associated culture collection (Microorganism and Gene Bank) enables physical access to strains.

The study presents ODFM, a web-based knowledgebase integrating genomes and multi-omic data for microorganisms across diverse fermented food ecosystems, together with built-in tools for similarity search, genome visualization, and ANI-based relatedness. It provides curated, downloadable datasets and facilitates evaluation of candidate fermentation starters and further functional studies. Future directions include continuous expansion of the archive, improved performance and features, incorporation of additional fermented food categories aligned with global consumer preferences, and broader accommodation of third-party data submissions to enhance community engagement and data sharing.

• Current data submission is centrally managed; external researcher self-registration is not yet enabled, potentially limiting rapid community-driven data growth.
• Fermentation phenotypes are highly influenced by environmental conditions and microbial community interactions, limiting direct translation from genomic potential to predictable fermentation performance.
• The database’s focus on fermentation-associated microbes, while a strength for domain specificity, may limit coverage for broader environmental or clinical comparisons.