A package containing clients of UTA.
Each acts as a data provider for hgvs, a library of sequence manipulation tools. Ensuring a data provider for hgvs allows users to normalize, validate, and map sequence variants, among other functionalities such as parsing and formatting that are included in the hgvs library.
Supported clients include:
Read the official cdot documentation here.
* = places where different transcripts have been returned due to cdot and uta being two different data sources/collectors
Data Provider Function | uta | cdot |
---|---|---|
get_seq | ||
get_acs_for_protein_seq | Not implemented (note in hgvs to drop this method) | |
get_gene_info | Includes datetime of when transcript was added | |
get_tx_exons | Includes the following additional fields: hgnc, tx_aseq, alt_aseq, tx_exon_set_id, alt_exon_set_id, tx_exon_id, alt_exon_id, exon_aln_id | |
get_tx_for_gene | * | * |
get_tx_for_region | ||
get_alignments_for_region | ||
get_tx_identity_info | ||
get_tx_info | ||
get_tx_mapping_options | ||
get_similar_transcripts | Not implemented (hgvs does not use this function) | |
get_pro_ac_for_tx_ac | ||
get_assembly_map |
$ pip install hgvs
$ pip install utaclients
In python:
>>> import hgvs
>>> import utaclients
>>> hdp = ...
Assign any of the following supported clients to your hgvs data provider.
>>> hdp = utaclients.uta.connect()
>>> hdp = utaclients.utarest.connect()
>>> hdp = utaclients.cdot.connect()
Then you can get started using all the functionalities of hgvs.
>>> hdp = utaclients.uta.connect()
>>> am = hgvs.assemblymapper.AssemblyMapper(hdp,
... assembly_name='GRCh37', alt_aln_method='splign',
... replace_reference=True)
$ make devready
$ source venv/bin/activate