Updating metadata

Dataset metadata, including links to external resources, can be added or updated using the blobtools add command. Given the focus on updating metadata, all examples use blobtools replace which is equivalent to blobtools add --replace.

Updating assembly/taxon information

Any values in the BlobDir dataset metadata can be updated by specifying --key path=value. In this context, path is a .-separated hierarchy of keys, e.g. assembly.accession.

~/blobtoolkit/blobtools2/blobtools replace \
     --key assembly.reference=doi:10.1038/ng.3495 \
     --key taxon.common_name=threadworm \
     ~/BTK_TUTORIAL/DATASETS/ASSEMBLY_NAME

Updating default plot axes

The default plot axes when a dataset is opened in the BlobToolKit Viewer, are defined in the plot section of the metadata. This will usually be:

• x – GC proportion
• y – First coverage field added (see Adding coverage)
• z – Length
• cat (categories) – Taxonomy from the first taxrule added at the phylum level (see Adding hits)

To update this, use the --key option as above:

~/blobtoolkit/blobtools2/blobtools replace \
     --key plot.x=length \
     --key plot.y=DRR008460_cov \
     --key plot.cat=bestsumorder_family \
     ~/BTK_TUTORIAL/DATASETS/ASSEMBLY_NAME

Adding links to external resources

Datasets can contain links to external resources, these are shown in the detail and table views of the BlobToolKit Viewer. Links are defined using URL templates that allow dataset specific terms to be included in the URL. The option to add new links (--link path=URL_template) uses path in the same way as for --key above for values in the metadata (e.g. assembly.accession), with additional values to allow links from individual contains or BLAST hits. A set of default links is available when starting the Viewer. When specified using JSON (in a file named default.json in the same directory as the datasets) these links can have custom titles. When added to the dataset, however, the link title is set using the last key in the path:

--link 'taxon.taxid=https://www.ebi.ac.uk/ena/browser/view/Taxon:{taxid}'
--link 'taxon.species=https://wikipedia.org/{species}'
--link 'assembly.bioproject=https://www.ebi.ac.uk/ena/browser/view/{bioproject}'

For comparison, more complex linking relationships can be specified in default.json or by directly editing the dataset meta.json file.

{
  ...
  "links": {
     "taxon": {
       "taxid" : {
         "ENA": "https://www.ebi.ac.uk/ena/browser/view/Taxon:{taxid}"
       },
       "species" : {
         "Wikipedia": "https://wikipedia.org/{species}"
       }
     },
     "blobtoolkit": {
       "commit" : {
         "Github": "{pipeline}/tree/{commit}"
       }
     },
     "assembly": {
       "biosample": {
         "ENA": "https://www.ebi.ac.uk/ena/browser/view/{biosample}"
       },
       "bioproject": {
         "ENA": "https://www.ebi.ac.uk/ena/browser/view/{bioproject}"
       }
     },
     "position": [
       {
         "patterns": [
           {
             "title": "NCBI RefSeq",
             "template": "https://www.ncbi.nlm.nih.gov/nuccore/{subject}",
             "regex": "^[NXW][A-Z]_.+"
           },
           {
             "title": "ENA",
             "template": "https://www.ebi.ac.uk/ena/browser/view/{subject}",
             "regex": "^.+$"
           }
         ]
       },
       {
         "UniProt": "https://www.uniprot.org/uniprot/{subject}"
       }
     ]
   }
   ...
 }

For links based on information in the metadata, the link will only be added if the link URL can be resolved. Adding the --skip-link-test flag will allow links to be added even if the URL cannot be resolved.