A Python interface to the Ensembl REST APIs. A whole world of biological data at your fingertips.
The Ensembl database contains reference biological data on almost any organism. Now it is easy to access this data programatically through their REST API.
The full list of endpoints for the Ensembl REST API endpoints along with endpoint-specific documentation can be found on their website.
This library also includes some utilities built on top of the APIs designed to ease working with them, including an AssemblyMapper class that helps in the conversion between different genome assemblies.
This project uses code from RESTEasy which made my life much easier. Thanks!
You can install from PyPI
$ pip install ensembl_rest
The library exports methods that point to each endpoint of the API, such as:
>>> import ensembl_rest
>>> ensembl_rest.symbol_lookup(
species='homo sapiens',
symbol='BRCA2'
)
{ 'species': 'human', 'object_type': 'Gene', 'description': 'BRCA2, DNA repair associated [Source:HGNC Symbol;Acc:HGNC:1101]', 'assembly_name': 'GRCh38', 'end': 32400266, ... ... ... 'seq_region_name': '13', 'strand': 1, 'id': 'ENSG00000139618', 'start': 32315474}
All the endpoints are listed on the API website. A quick lookup of the methods can be obtained by calling help on the module:
>>> help(ensembl_rest)
If you want to use an endpoint from the ones enlisted in the API website,
say GET lookup/symbol/:species/:symbol
,
then the name of the corresponding method is in the endpoint documentation URL,
in this case, the documentation links to
http://rest.ensembl.org/documentation/info/symbol_lookup so the
corresponding method name is symbol_lookup
.
>>> help(ensembl_rest.symbol_lookup)
Help on function symbol_lookup in module ensembl_rest: symbol_lookup(*args, **kwargs) Lookup ``GET lookup/symbol/:species/:symbol`` Find the species and database for a symbol in a linked external database **Parameters** - Required: + **Name**: species + *Type*: String + *Description*: Species name/alias + *Default*: - + *Example Values*: homo_sapiens, human ... ... - Optional: + **Name**: expand + *Type*: Boolean(0,1) + *Description*: Expands the search to include any connected features. e.g. If the object is a gene, its transcripts, translations and exons will be returned as well. ... ... **Resource info** - **Methods**: GET - **Response formats**: json, xml, jsonp **More info** https://rest.ensembl.org/documentation/info/symbol_lookup
We can see from the resource string GET lookup/symbol/:species/:symbol
that
this method contains 2 parameters called species and symbol, so we can call the
method in the following way:
>>> ensembl_rest.symbol_lookup(
species='homo sapiens',
symbol='TP53'
)
# Or like this...
>>> ensembl_rest.symbol_lookup('homo sapiens', 'TP53')
{'source': 'ensembl_havana', 'object_type': 'Gene', 'logic_name': 'ensembl_havana_gene', ... ... ... 'start': 32315474}
One can provide optional parameters with the params
keyword (the specific parameters to pass depend on the specific endpoint,
the official endpoints documentation can be found here)_:
# Fetch also exons, transcripts, etc...
>>> ensembl_rest.symbol_lookup('human', 'BRCA2',
params={'expand':True})
{'source': 'ensembl_havana', 'seq_region_name': '13', 'Transcript': [{'source': 'ensembl_havana', 'object_type': 'Transcript', 'logic_name': 'ensembl_havana_transcript', 'Exon': [{'object_type': 'Exon', 'version': 4, 'species': 'human', 'assembly_name': 'GRCh38', ... ... ... 'biotype': 'protein_coding', 'start': 32315474}
The parameters for the POST endpoints are also provided via the params
keyword , such as in the next example:
>>> ensembl_rest.symbol_post(species='human',
params={'symbols': ["BRCA2",
"TP53",
"BRAF" ]})
{ "BRCA2": { "source": "ensembl_havana", "object_type": "Gene", "logic_name": "ensembl_havana_gene", "description": "BRCA2, DNA repair associated [Source:HGNC Symbol;Acc:HGNC:1101]", ... ... }, "TP53": { ... ... }. "BRAF": { ... ... "strand": -1, "id": "ENSG00000157764", "start": 140719327 } }
Another common usage is to fetch sequences of known genes:
>>> ensembl_rest.sequence_id('ENSG00000157764')
{'desc': 'chromosome:GRCh38:7:140719327:140924928:-1', 'query': 'ENSG00000157764', 'version': 13, 'id': 'ENSG00000157764', 'seq': 'TTCCCCCAATCCCCTCAGGCTCGG...ATTGACTGCATGGAGAAGTCTTCA', 'molecule': 'dna'}
if you want it in FASTA, you can modify the headers
:
>>> ensembl_rest.sequence_id(
'ENSG00000157764',
headers={'content-type': 'text/x-fasta'})
>ENSG00000157764.13 chromosome:GRCh38:7:140719327:140924928:-1 TTCCCCCAATCCCCTCAGGCTCGGCTGCGCCCGGGGCCGCGGGCCGGTACCTGAGGTGGC CCAGGCGCCCTCCGCCCGCGGCGCCGCCCGGGCCGCTCCTCCCCGCGCCCCCCGCGCCCC CCGCTCCTCCGCCTCCGCCTCCGCCTCCGCCTCCCCCAGCTCTCCGCCTCCCTTCCCCCT ...
Notice that, if left unchanged, the methods ask for data in dictionary (JSON) format so that they are easy to use. If the response cannot be decoded as such, then it is returned as plain text, such as the above.
You can also map betweeen assemblies...
>>> ensembl_rest.assembly_map(species='human',
asm_one='GRCh37',
region='X:1000000..1000100:1',
asm_two='GRCh38')
# Or...
>>> region_str = ensembl_rest.region_str(chrom='X',
start=1000000,
end=1000100)
>>> ensembl_rest.assembly_map(species='human',
asm_one='GRCh37',
region=region_str,
asm_two='GRCh38')
{'mappings': [{'original': {'seq_region_name': 'X', 'strand': 1, 'coord_system': 'chromosome', 'end': 1000100, 'start': 1000000, 'assembly': 'GRCh37'}, 'mapped': {'seq_region_name': 'X', 'strand': 1, 'coord_system': 'chromosome', 'end': 1039365, 'start': 1039265, 'assembly': 'GRCh38'}}]}
The above problem (mapping from one assembly to another) is so frequent that
the library provides a specialized class AssemblyMapper
to efficiently
mapping large amounts of regions between assemblies. This class avoids the
time-consuming task of making a web request every time a mapping is needed by
fetching the mapping of the whole assembly right from the instantiation. This
is a time-consuming operation by itself, but it pays off when one has to
transform repeatedly betweeen assemblies.:
>>> mapper = ensembl_rest.AssemblyMapper( species='human', from_assembly='GRCh37', to_assembly='GRCh38' ) >>> mapper.map(chrom='1', pos=1000000) 1064620
You can also find orthologs, paralogs and gene tree information, along with variation data and basically everything Ensembl has to offer.
If you want to instantiate your own client, you can do it by using the
ensembl_rest.EnsemblClient
class, this class is the one that contains all
the endpoint methods.
>>> client = ensembl_rest.EnsemblClient()
>>> client.symbol_lookup('homo sapiens', 'TP53')
{'source': 'ensembl_havana', 'object_type': 'Gene', 'logic_name': 'ensembl_havana_gene', 'version': 14, 'species': 'human', ... ... ...}
Finally, the library exposes the class ensembl_rest.HTTPError
that allows to
handle errors in the requests. An example of it's utility is when using the
GET genetree/member/symbol/:species/:symbol
endpoint to query for gene trees
in order to find ortholog and paralog proteins and genes. This endpoint returns
an HTTP error when a gene tree is not found with code 400 and the error message
Lookup found nothing
. We can use this information to detect the error
and handle it, or to simply ignore it if we expected it:
for gene in ['TP53', 'rare-new-gene', 'BRCA2']:
try:
gene_tree = ensembl_rest.genetree_member_symbol(
species='human',
symbol=gene,
params={'prune_species': 'human'}
)
# Assuming we have a function to extract the paralogs
paralogs = extract_paralogs(gene_tree['tree'])
print(paralogs)
# Handle the case when there's no gene tree
except ensembl_rest.HTTPError as err:
error_code = err.response.status_code
error_message = err.response.json()['error']
if (error_code == 400) \
and ('Lookup found nothing' in error_message):
# Skip the gene with no data
pass
else:
# The exception was caused by another problem
# Raise the exception again
raise
Author: Ad115 - Github – a.garcia230395@gmail.com
Project pages: Docs - @GitHub - @PyPI
Distributed under the MIT license. See LICENSE for more information.
- Check for open issues or open a fresh issue to start a discussion around a feature idea or a bug.
- Fork the repository on GitHub to start making your changes to a feature branch, derived from the master branch.
- Write a test which shows that the bug was fixed or that the feature works as expected.
- Send a pull request and bug the maintainer until it gets merged and published.