def test_get_signal_peptide():
# This is a valid entry local to the resource file
sp = uniprot_client.get_signal_peptide('P00533')
assert sp.begin == 1, sp
assert sp.end == 24, sp
# This one requires a web lookup
sp = uniprot_client.get_signal_peptide('P00534')
assert sp is None, sp
# This one errors when doing web lookup
sp = uniprot_client.get_signal_peptide('Q9H7H1')
assert sp is None, sp
def test_get_signal_peptide():
# This is a valid entry local to the resource file
bp, ep = uniprot_client.get_signal_peptide('P00533')
assert bp == 1, bp
assert ep == 24, ep
# This one requires a web lookup
bp, ep = uniprot_client.get_signal_peptide('P00534')
assert bp is None, bp
assert ep is None, ep
# This one errors when doing web lookup
bp, ep = uniprot_client.get_signal_peptide('Q9H7H1')
assert bp is None, bp
assert ep is None, ep
def map_to_human_ref(self, prot_id, prot_ns, residue, position,
do_methionine_offset=True,
do_orthology_mapping=True,
do_isoform_mapping=True):
"""Check an agent for invalid sites and look for mappings.
Look up each modification site on the agent in Uniprot and then the
site map.
Parameters
----------
prot_id : str
A Uniprot ID or HGNC gene symbol for the protein.
prot_ns : str
One of 'uniprot' or 'hgnc' indicating the type of ID given.
residue : str
Residue to map on the protein to check for validity and map.
position : str
Position of the residue to check for validity and map.
do_methionine_offset : boolean
Whether to check for off-by-one errors in site position (possibly)
attributable to site numbering from mature proteins after
cleavage of the initial methionine. If True, checks the reference
sequence for a known modification at 1 site position greater
than the given one; if there exists such a site, creates the
mapping. Default is True.
do_orthology_mapping : boolean
Whether to check sequence positions for known modification sites
in mouse or rat sequences (based on PhosphoSitePlus data). If a
mouse/rat site is found that is linked to a site in the human
reference sequence, a mapping is created. Default is True.
do_isoform_mapping : boolean
Whether to check sequence positions for known modifications
in other human isoforms of the protein (based on PhosphoSitePlus
data). If a site is found that is linked to a site in the human
reference sequence, a mapping is created. Default is True.
Returns
-------
MappedSite
The MappedSite object gives information on results of mapping the
site. See :py:class:`protmapper.api.MappedSite` documentation for
details.
"""
# Check the protein ID and namespace
if prot_id is None:
raise ValueError("prot_id must not be None.")
if prot_ns not in ('uniprot', 'hgnc'):
raise ValueError("prot_ns must be either 'uniprot' or 'hgnc' (for "
"HGNC symbols)")
# Get Uniprot ID and gene name
up_id = _get_uniprot_id(prot_id, prot_ns)
# If an HGNC ID was given and the uniprot entry is not found, flag
# as error
if up_id is None:
assert prot_ns == 'hgnc' and prot_id is not None
return MappedSite(None, None, residue, position,
gene_name=prot_id, error_code='NO_UNIPROT_ID')
# Make sure the sites are proper amino acids/positions
try:
valid_res, valid_pos = _validate_site(residue, position)
except InvalidSiteException as ex:
return MappedSite(up_id, None, residue, position,
error_code='INVALID_SITE',
description=str(ex))
# Get the gene name from Uniprot
gene_name = uniprot_client.get_gene_name(up_id, web_fallback=False)
site_key = (up_id, residue, position)
# First, check the cache to potentially avoid a costly sequence
# lookup
cached_site = self._cache.get(site_key)
if cached_site is not None:
return cached_site
# If not cached, continue
# Look up the residue/position in uniprot
try:
site_valid = uniprot_client.verify_location(up_id, residue,
position)
error_code = None
except HTTPError as ex:
if ex.response.status_code == 404:
error_code = 'UNIPROT_HTTP_NOT_FOUND'
else:
error_code = 'UNIPROT_HTTP_OTHER'
except Exception as ex:
error_code = 'UNIPROT_OTHER'
logger.error(ex)
if error_code:
# Set error_code; valid will set to None, not True/False
mapped_site = MappedSite(up_id, None, residue, position,
error_code=error_code)
return mapped_site
# It's a valid site
if site_valid:
mapped_site = MappedSite(up_id, True, residue, position,
description='VALID',
gene_name=gene_name)
self._cache[site_key] = mapped_site
return mapped_site
# If it's not a valid site, check the site map first
curated_site = self.site_map.get(site_key, None)
# Manually mapped in the site map
if curated_site is not None:
mapped_res, mapped_pos, description = curated_site
mapped_site = MappedSite(up_id, False, residue, position,
mapped_id=up_id,
mapped_res=mapped_res,
mapped_pos=mapped_pos,
description=description,
gene_name=gene_name)
self._cache[site_key] = mapped_site
return mapped_site
# There is no manual mapping, next we try to see if UniProt
# reports a signal peptide that could be responsible for the position
# being shifted
signal_peptide = uniprot_client.get_signal_peptide(up_id, False)
# If there is valid signal peptide information from UniProt
if signal_peptide and signal_peptide.begin == 1 and \
signal_peptide.end is not None:
offset_pos = str(int(position) + signal_peptide.end)
# Check to see if the offset position is known to be phosphorylated
mapped_site = self.get_psp_mapping(
up_id, up_id, gene_name, residue, position,
offset_pos, 'SIGNAL_PEPTIDE_REMOVED')
if mapped_site:
return mapped_site
# ...there's no manually curated site or signal peptide, so do mapping
# via PhosphoSite if the data is available:
human_prot = uniprot_client.is_human(up_id)
if phosphosite_client.has_data():
# First, look for other entries in phosphosite for this protein
# where this sequence position is legit (i.e., other isoforms)
if do_isoform_mapping and up_id and human_prot:
mapped_site = self.get_psp_mapping(
up_id, up_id, gene_name, residue, position, position,
'INFERRED_ALTERNATIVE_ISOFORM')
if mapped_site:
return mapped_site
# Try looking for rat or mouse sites
if do_orthology_mapping and up_id and human_prot:
# Get the mouse ID for this protein
up_mouse = uniprot_client.get_mouse_id(up_id)
# Get mouse sequence
mapped_site = self.get_psp_mapping(
up_id, up_mouse, gene_name, residue,
position, position, 'INFERRED_MOUSE_SITE')
if mapped_site:
return mapped_site
# Try the rat sequence
up_rat = uniprot_client.get_rat_id(up_id)
mapped_site = self.get_psp_mapping(
up_id, up_rat, gene_name, residue, position,
position, 'INFERRED_RAT_SITE')
if mapped_site:
return mapped_site
# Check for methionine offset (off by one)
if do_methionine_offset and up_id and human_prot:
offset_pos = str(int(position) + 1)
mapped_site = self.get_psp_mapping(
up_id, up_id, gene_name, residue, position,
offset_pos, 'INFERRED_METHIONINE_CLEAVAGE')
if mapped_site:
return mapped_site
# If we've gotten here, the entry is 1) not in the site map, and
# 2) we either don't have PSP data or no mapping was found using PSP
mapped_site = MappedSite(up_id, False, residue, position,
description='NO_MAPPING_FOUND',
gene_name=gene_name)
self._cache[site_key] = mapped_site
return mapped_site