Folder structure

folder_structure
├── version.json
└── organisms
     ├── ORGANISM1
     ├── ...
     └── EXAMPLE1234
         ├── organism.json
         ├── organism.md (optional)
         └── genomes
            ├── EXAMPLE1234-1-1.1
            └── EXAMPLE1234-2-1.1
              ├── genome.json
              ├── genome.md (optional)
              ├── assembly.fna
              ├── protein.faa
              ├── genbank.gbk
              ├── generalfeatureformat.gff
              ├── annotations.ko (optional)
              └── annotations.eggnog (optional)

A shortcut: opengenomebrowser-tools

The python library opengenomebrowser-tools can create both the folder structure and the metadata files automatically! The relevant commands are init_folder_structure and import_genome.

Versioning

From time to time, changes are made to the OpenGenomeBrowser folder structure. This is reflected in version.json. The folder structure and the OpenGenomeBrowser code must always be compatible. If they are not, OpenGenomeBrowser will not start and give an error message. In this case, the relevant updates must be applied to the folder structure. The command update_folder_structure in the library opengenomebrowser-tools should make this very easy. Example usage:

export FOLDER_STRUCTURE=/path/to/folder_structure
update_folder_structure 1_to_2

Rules

• genome names must start with the name of the corresponding organism
  • we suggest to use this suffix format: organism-isolate-assembly.annotation
  • example: organism=EXAMPLE1234 -> EXAMPLE1234-2-1.1
• gene locus tags must start with the genome identifier, followed by an underline (_)
  • example: EXAMPLE1234-2-1.1_000001

Metadata files

Notes:

• paths are relative to the genome folder, i.e. it’s fine to have files in a subfolders of the genome folder
• date format: "%Y-%m-%d", i.e. 2000-12-31

organism.json:

Click here to expand.

In the example below, all mandatory fields contain dummy data and all non-mandatory fields have null-input.

{
  "name": "EXAMPLE1234",
  "alternative_name": null,
  "taxid": null,
  "restricted": false,
  "tags": [],
  "representative": "EXAMPLE1234-2-1.1"
}

genome.json:

Click here to expand.

In the example below, all mandatory fields contain dummy data and all non-mandatory fields have null-input.

{
  "identifier": "EXAMPLE1234-2-1.1",
  "contaminated": false,
  "old_identifier": null,
  "isolation_date": null,
  "env_broad_scale": [],
  "env_local_scale": [],
  "env_medium": [],
  "growth_condition": null,
  "geographical_coordinates": null,
  "geographical_name": null,
  "library_preparation": null,
  "sequencing_tech": null,
  "sequencing_tech_version": null,
  "sequencing_date": null,
  "sequencing_coverage": null,
  "read_length": null,
  "assembly_tool": null,
  "assembly_version": null,
  "assembly_date": null,
  "nr_replicons": null,
  "cds_tool": null,
  "cds_tool_date": null,
  "cds_tool_version": null,
  "cds_tool_faa_file": "protein.faa",
  "cds_tool_ffn_file": "nucleotide_cds.ffn",
  "cds_tool_gbk_file": "genbank.gbk",
  "cds_tool_gff_file": "generalfeatureformat.gff",
  "cds_tool_sqn_file": null,
  "assembly_fasta_file": "assembly.fna",
  "custom_annotations": [],
  "BUSCO": {},
  "COG": {},
  "bioproject_accession": null,
  "biosample_accession": null,
  "genome_accession": null,
  "literature_references": [],
  "custom_tables": {},
  "tags": []
}

BUSCO format:

"BUSCO": {
  "C": 0,
  "D": 0,
  "F": 0,
  "M": 0,
  "S": 0,
  "T": 0,
  "dataset": "lactobacillales_odb10",
  "dataset_creation_date": "2000-12-31"
}

The keys dataset and dataset_creation_date are optional.

opengenomebroser-tools can automatically parse this information from busco-output-files.

COG format:

"COG": {
  "J": 152.5,
  "A": 2,
  "K": 29.3,
  "L": 65.833
}

These are allowed COG categories: JAKLBDYVTMNZWUOXCGEFHIPQRS. If a category is omitted, it is assumed to be zero. opengenomebroser-tools can automatically parse this information from eggnog-files.

Custom annotations format - type eggnog:

For files created by eggNOG v2-v2.1.1:

"custom_annotations": [
  {
    "date": "0000-00-00",
    "file": "annotations.eggnog",
    "type": "__replace_this__"
  }
]

If your file was created by eggNOG

• v2-v2.1.1: type must be eggnog
• v2.1.2+: type must be eggnog-2.1.2

The file annotations.eggnog must be an eggNOG-mapper v2 output file. (Usual filename: query_seqs.fa.emapper.annotations) opengenomebroser-tools can automatically detect these files.

Custom annotations format:

Click here for more information about custom annotations. Click here to learn how to format the file.

"custom_annotations": [
  {
    "date": "0000-00-00",
    "file": "annotations.ko",
    "type": "KG"
  }
]

opengenomebroser-tools can automatically detect these files.

Custom tables format:

"custom_tables": {
  "Table 1 Title": {
    "index_col": "taxid",
    "taxid_cols": ["taxid"],
    "rows": [
      {
        "description": "Fakebacillus bullshitingis organism 42 16S ribosomal RNA, partial sequence",
        "taxid": 0,
        "evalue": 0.0
      }
    ]
  }
}

Alternatively, the markdown files (organism.md or genome.md) could be used to add tables.

Literature references format:

"literature_references": [
  {
    "url": "https://pubmed.ncbi.nlm.nih.gov/1234/",
    "name": "Name of paper"
  }
]

organism.md and genome.md

Click here to expand.

In addition to the more restrictive .json files, a .md file can optionally be added to each organism and genome.

The Markdown format is very flexible and allows users to add text, links and tables to these pages. Users with staff permissions can edit these Markdown files directly through OpenGenomeBrowser via the user button on the these pages.

• organism.md will be integrated on the main page of each organism (/organism/<ORGANISM>) and genome (/genome/<GENOME>)
• genome.md will be integrated on the main page of each genome (/genome/<GENOME>)