def _get_curie_and_type_from_id(variant_id):
"""
Given a variant id, our best guess at its curie and type (snp, haplotype, etc)
'None' will be used for both curie and type for IDs that we can't process
# 2019-May three snp-id have ' e' or ' a' appended. note space.
# examples: 'rs2440154 e-A' and 'rs2440154 e'
# including the suffix in the url is a web noop but breaks rdflib
:param variant_id:
:return:
"""
curie = None
variant_type = None
# remove space before hyphens
variant_id = re.sub(r' -', '-', variant_id).strip()
if re.search(r' x ', variant_id) or re.search(r',', variant_id):
# TODO deal with rs1234 x rs234... (haplotypes?)
LOG.warning("Cannot parse variant groups of this format: %s",
variant_id)
elif re.search(r';', variant_id):
curie = ':haplotype_' + Source.hash_id(
variant_id) # deliberate 404
variant_type = "haplotype"
elif variant_id[:2] == 'rs':
# remove whitespace from errant id, rs6194 5053-?
curie = 'dbSNP:' + variant_id.split('-')[0].replace(' ', '')
# curie = re.sub(r'-.*$', '', curie).strip()
variant_type = "snp"
# remove the alteration
elif variant_id[:3] == 'kgp':
# http://www.1000genomes.org/faq/what-are-kgp-identifiers
curie = 'GWAS:' + variant_id.split('-')[0]
variant_type = "snp"
elif variant_id[:3] == 'chr':
# like: chr10:106180121-G
variant_id = re.sub(r'-?', '-N', variant_id)
variant_id = re.sub(r' ', '', variant_id)
# going to hate myself but ...
# moving this from a broken base node to yet another blank node
# It had produced this monstrocity with the embedded quote
# :gwas--Nc-Nh-Nr-N1-N1-N--N1-N0-N2-N7-N5-N1-N1-N0-N2-N"-N?-N
curie = Source.make_id('gwas-' + re.sub(r':', '-', variant_id),
'_')
variant_type = "snp"
elif variant_id.strip() == '':
pass
else:
LOG.warning("There's a snp id i can't manage: %s", variant_id)
return curie, variant_type
def make_reagent_targeted_gene_id(gene_id, reagent_id):
return Source.make_id('-'.join((gene_id, reagent_id)), '_')
