Basics

django-biodas uses django-tastypie as the basis for generating a xml response that follows the biodas specification.

Adding a Django Model

If you already have a django ORM for biological data, it is straigtforward to generate a biodas compliant api using django-biodas.

1. Add tastypie and biodas to your ‘’INSTALLED_APPS’‘.

2. Create an api diretory in your app with an empty ‘__init__.py’‘.

3. In the api directory, create a ‘’<my_app>/api/resource.py’’ file and place the following in it:

   from biodas import DasModelResource
   from yourapp.models import YourModel
   
   class YourResource(DasModelResource):
       class Meta:
           resource_name ='yourdata'
           queryset = YourModel.objects.all()
   
           version = '36'
           authority = 'NCBI'
   

4. Register the resource with the DAS api in ‘’urls.py’‘

   from biodas import DAS
   from api.resource import YourResource
   
   api = DAS()
   api.register(YourResource())
   
   urlpatterns = patterns('',
     (r'^api/', include(api.urls)),
     )
   

Doing so generates a api that is accessible from this url::

/api/yourdata

Querying over features::

/api/yourdata/features?segment=chr1:20,60

Excluding Fields from the return can be done much like in django-tastypie:

class YourResource(DasModelResource):
    class Meta:
        resource_name ='yourdata'
        queryset = YourModel.objects.all()

        version = '36'
        authority = 'NCBI'
        excludes = ['CHROM', 'Extraneous_fields']

note:

By default, fields such as chromosome, region, or segement id are left out
of the response as to meet the biodas 1.6 specification.

Examples of models resources can be found in tests/core/tests/api.py

Generating a resource from a file

Additionally django-biodas makes it really easy to generate an api from a file such as a BAM, BIGWIG, VCF, BED, GFF or BIGBED file:

from biodas import DasFileResource

class FileResource(DasFileResource):
    class Meta:
        filename = 'my.gff'
        resource_name = 'mygff'

biodas attempts to intelligently parse the file format. You can manually set the file format using::

class Meta:
    filetype = 'gff'

Note unindexed files such as BEDFiles will have long query times.

You can also generate a reference server in addition to a feature server.

JSON based responses

Although the xml response is the default response for biodas servers, biodas has the option to return JSON responses. To serialize features in json, simply set json to True in the meta file. The specification for the JSON response can be found :TODO

Registering the server with the DAS registry

Not yet implemented