What it will do:

• Show all databases of a biomart server

• Show all datasets of a biomart database

• Show attributes and filters of a biomart dataset

• Run your query formatted as a Python dict and return the Biomart response as TSV format.

What it won’t do:

• Process and return the results as JSON,XML,etc.

Usage

Import Biomart module

from biomart import BiomartServer

Connect to a Biomart Server

server = BiomartServer( "http://www.biomart.org/biomart" )

# if you are behind a proxy
import os
server.http_proxy = os.environ.get('http_proxy', 'http://my_http_proxy.org')

# set verbose to True to get some messages
server.verbose = True

Interact with the biomart server

# show server databases
server.show_databases() # uses pprint behind the scenes

# show server datasets
server.show_datasets() # uses pprint behind the scenes

# use the 'uniprot' dataset
uniprot = server.datasets['uniprot']

# show all available filters and attributes of the 'uniprot' dataset
uniprot.show_filters()  # uses pprint
uniprot.show_attributes()  # uses pprint

Run a search

# run a search with the default attributes - equivalent to hitting "Results" on the web interface.
# this will return a lot of data.
response = uniprot.search()
response = uniprot.search( header = 1 ) # if you need the columns header

# response format is TSV
for line in response.iter_lines():
  line = line.decode('utf-8')
  print(line.split("\t"))

# run a count - equivalent to hitting "Count" on the web interface
response = uniprot.count()
print(response.text)

# run a search with custom filters and default attributes.
response = uniprot.search({
  'filters': {
      'accession': 'Q9FMA1'
  }
}, header = 1 )

response = uniprot.search({
  'filters': {
      'accession': ['Q9FMA1', 'Q8LFJ9']  # ID-list specified accessions
  }
}, header = 1 )

# run a search with custom filters and attributes (no header)
response = uniprot.search({
  'filters': {
      'accession': ['Q9FMA1', 'Q8LFJ9']
  },
  'attributes': [
      'accession', 'protein_name'
  ]
})

Shortcut function: connect directly to a biomart dataset This is short in code but it might be long in time since the module needs to fetch all server’s databases to find your dataset.

from biomart import BiomartDataset

interpro = BiomartDataset( "http://www.biomart.org/biomart", name = 'entry' )

response = interpro.search({
  'filters': { 'entry_id': 'IPR027603' },
  'attributes': [ 'entry_name', 'abstract' ]
})