Skip to content

Commit

Permalink
cross scv links exist
Browse files Browse the repository at this point in the history
  • Loading branch information
@TomConlin
TomConlin committed May 20, 2016
1 parent c249fd5 commit 3b0fc3e
Showing 1 changed file with 51 additions and 45 deletions.
96 changes: 51 additions & 45 deletions dipper/sources/ClinVarXML_alpha.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -12,10 +12,14 @@
import sys
import gzip
import hashlib
import logging
import argparse
import xml.etree.ElementTree as ET
# import Requests


logger = logging.getLogger(__name__)

# from dipper import curie_map # hangs on to stale data?


Expand All @@ -27,15 +31,15 @@

# The name of the ingest we are doing
IPATH = re.split(r'/', os.path.realpath(__file__))
(INAME, dotpy) = re.split(r'\.', IPATH[-1].lower())
(INAME, DOTPY) = re.split(r'\.', IPATH[-1].lower())
RPATH = '/' + '/'.join(IPATH[1:-3])
FILES = {'f1': 'ClinVarFullRelease_00-latest.xml.gz'}

# I am not positive allowing the slash is ligit
CURIERE = re.compile(r'^.*:[A-Za-z0-9_][A-Za-z0-9_./]*[A-Za-z0-9_]$')


ENIGMA = 'https://submit.ncbi.nlm.nih.gov/ft/byid/hxnfuuxx/enigma_rules_2015-03-26.pdf'
ENIGMA = \
'https://submit.ncbi.nlm.nih.gov/ft/byid/hxnfuuxx/enigma_rules_2015-03-26.pdf'

# handle arguments for IO
argparser = argparse.ArgumentParser()
Expand Down Expand Up @@ -82,7 +86,7 @@
'foaf': 'http://xmlns.com/foaf/0.1/',
'_': 'https://monarchinitiave.org/.well-known/genid/',
'BFO': 'http://purl.obolibrary.org/obo/BFO_',
'ERO' : 'http://purl.obolibrary.org/obo/ERO_',
'ERO': 'http://purl.obolibrary.org/obo/ERO_',
'GENO': 'http://purl.obolibrary.org/obo/GENO_',
'GO': 'http://purl.obolibrary.org/obo/GO_',
'RO': 'http://purl.obolibrary.org/obo/RO_',
Expand All @@ -93,9 +97,9 @@
'NCBITaxon': 'http://purl.obolibrary.org/obo/NCBITaxon_',
'NCBIGene': 'http://www.ncbi.nlm.nih.gov/gene/',
'MmusDv': 'http://purl.obolibrary.org/obo/MmusDv_',
'OBAN': 'http://purl.org/oban/',
'owl': 'http://www.w3.org/2002/07/owl#',
'OBO': 'http://purl.obolibrary.org/obo/',
'OIO': 'http://www.geneontology.org/formats/oboInOwl#',
'SEPIO': 'http://purl.obolibrary.org/obo/SEPIO_',
'ClinVar': 'http://www.ncbi.nlm.nih.gov/clinvar/',
'ClinVarVariant': 'http://www.ncbi.nlm.nih.gov/clinvar/variation/',
Expand Down Expand Up @@ -128,7 +132,7 @@ def make_spo(sub, prd, obj):

if match is not None and objcuri in CURIEMAP:
objt = '<' + CURIEMAP[objcuri] + objid + '>'
elif(obj.isnumeric()):
elif obj.isnumeric():
objt = obj
else:
objt = '"' + obj.strip('"') + '"'
Expand All @@ -140,7 +144,7 @@ def make_spo(sub, prd, obj):
def scv_link(scv_sig):
'''
Creates links between SCV based on their pathonnicty significancce
# GENO:0000840 - GENO:0000840 --> equivalent_to SEPIO:0000098
# GENO:0000841 - GENO:0000841 --> equivalent_to SEPIO:0000098
# GENO:0000843 - GENO:0000843 --> equivalent_to SEPIO:0000098
Expand All @@ -158,6 +162,7 @@ def scv_link(scv_sig):
'GENO:0000841': 2, # likely pathogenic
'GENO:0000844': 4, # likely benign
'GENO:0000843': 8} # benign

lnk = {
0: 'SEPIO:0000098',
1: 'SEPIO:0000099',
Expand All @@ -166,14 +171,15 @@ def scv_link(scv_sig):
4: 'SEPIO:0000099',
6: 'SEPIO:0000101',
7: 'SEPIO:0000100'}

for scv_a in scv_sig.keys():
scv_av = scv_sig[scv_a]
scv_sig.remove(scv_a)
keys = sorted(scv_sig.keys())
for scv_a in keys:
scv_av = scv_sig.pop(scv_a)
for scv_b in scv_sig.keys():
link = lnk[abs(sig[scv_av] - sig[scv_sig[scv_b]])]
print(make_spo(scv_a, link, scv_b))
print(make_spo(scv_b, link, scv_a))
if scv_av in sig and scv_sig[scv_b] in sig:
link = lnk[abs(sig[scv_av] - sig[scv_sig[scv_b]])]
print(make_spo(scv_a, link, scv_b))
print(make_spo(scv_b, link, scv_a))


################################################################
# CONSTANTS once at the beginning (would be better not at all).
Expand Down Expand Up @@ -208,7 +214,6 @@ def scv_link(scv_sig):
print(make_spo('OBO:RO_0003303', 'rdf:type', 'owl:ObjectProperty'))
print(make_spo('OBO:GENO_0000418', 'rdf:type', 'owl:ObjectProperty'))


# larval stage term mapping file
# will want namespace I expect.
# strips comments and blank lines
Expand All @@ -227,19 +232,18 @@ def scv_link(scv_sig):
TREE = ET.parse(fh)
ReleaseSet = TREE.getroot()
if ReleaseSet.get('Type') != 'full':
print("Not a full release", file=sys.stderr)
logger.warning('Not a full release')
sys.exit(-1)

rs_dated = ReleaseSet.get('Dated') # "2016-03-01 (date_last_seen)

for ClinVarSet in ReleaseSet.findall('ClinVarSet[RecordStatus]'):
if ClinVarSet.find('RecordStatus').text != 'current':
print(
ClinVarSet.get('ID') + " <is not current as of> " +
rs_dated + ' .', file=sys.stderr)
logger.warning(
ClinVarSet.get('ID') + " is not current as of " + rs_dated)
continue # or break?

# collect svc significance calls
# collect svc significance calls within a rcv
pathocalls = {}

# There is only one RCV per ClinVarSet
Expand All @@ -257,8 +261,9 @@ def scv_link(scv_sig):

# I do not expect we care as we shouldn't keep the RCV.
if RCVAssertion.find('RecordStatus').text != 'current':
print(
rcv_acc + " <is not current on> " + rs_dated, file=sys.stderr)
logger.warning(
rcv_acc + " <is not current on> " + rs_dated)
continue

# Child elements
#
Expand Down Expand Up @@ -288,18 +293,18 @@ def scv_link(scv_sig):
RCV_MeasureSet.findall('Measure'):
rcv_variant_type = TT.get(RCV_Measure.get('Type'))
if rcv_variant_type is None:
print(
logger.warning(
rcv_acc + " UNKNOWN VARIANT TYPE " +
RCV_Measure.get('Type').text, file=sys.stderr)
RCV_Measure.get('Type').text)
continue

RCV_VariantName = RCV_Measure.find(
'Name/ElementValue[@Type="Preferred"]')
if RCV_VariantName is not None:
rcv_variant_label = RCV_VariantName.text
else:
print(
rcv_acc + " VARIANT MISSING LABEL", file=sys.stderr)
logger.warning(
rcv_acc + " VARIANT MISSING LABEL")

# /RCV/MeasureSet/Measure/Name/ElementValue/[@Type="Preferred"]
#######################################################################
Expand Down Expand Up @@ -327,7 +332,7 @@ def scv_link(scv_sig):
rcv_disease_label = RCV_TraitName.text
# print(rcv_acc + ' ' + rcv_disease_label)
else:
print(rcv_acc + " MISSING DISEASE NAME ", file=sys.stderr)
logger.warning(rcv_acc + " MISSING DISEASE NAME")

# Prioritize OMIM
for RCV_Trait in RCV_TraitSet.findall('Trait[@Type="Disease"]'):
Expand Down Expand Up @@ -368,10 +373,10 @@ def scv_link(scv_sig):
for RCV_Trait in\
RCV_TraitSet.findall('Trait[@Type="Disease"]'):
for RCV_TraitXRef in RCV_Trait.findall('XRef'):
print(
logger.warning(
rcv_acc + " UNKNOWN DISEASE DB:\t" +
RCV_TraitXRef.get('DB') + ":" +
RCV_TraitXRef.get('ID'), file=sys.stderr)
RCV_TraitXRef.get('ID'))
# 82372 MedGen
# 58 EFO
# 1 Human Phenotype Ontology
Expand All @@ -382,7 +387,7 @@ def scv_link(scv_sig):
if rcv_disease_db is None or rcv_disease_id is None or \
rcv_disease_label is None or rcv_variant_id is None or \
rcv_variant_type is None or rcv_variant_label is None:
print(rcv_acc + " RCV IS WONKY, BYEBYE", file=sys.stderr)
logger.warning(rcv_acc + " RCV IS WONKY, BYEBYE")
continue

rcv_disease_curi = rcv_disease_db + ':' + rcv_disease_id
Expand All @@ -391,11 +396,9 @@ def scv_link(scv_sig):
# Descend into each SCV grouped with the current RCV
#######################################################################

# keep a collection of an RCV's associations and their patho call
# when the collection is complete, i.e when pathocalls isn't empty here
# keep a collection of a SCV's associations and patho significance call
# when the collection is complete,
# interlink based on patho call
if len(pathocalls) > 0:
link_scv(pathocalls)

pathocalls = {}

Expand Down Expand Up @@ -518,14 +521,14 @@ def scv_link(scv_sig):
'AttributeSet/Attribute[@Type="AssertionMethod"]')
if SCV_Attribute is not None:
scv_assert_method = SCV_Attribute.text
# this string needs to be mapped to a <sepio:100...n> class curie
# this string needs to be mapped to a <sepio:100...n> curie
if scv_assert_method in TT:
scv_assert_id = TT[scv_assert_method]
# TRIPLES specified_by
# <:_assertion_id><SEPIO:0000041><scv_att_id>
print(make_spo(
_assertion_id, 'SEPIO:0000041', scv_assert_id))
# <scv_assert_id> <class?> <SEPIO:0000037>
# <scv_assert_id> <class?> <SEPIO:0000037>

# <scv_att_id><rdf:type><SEPIO:1000001>
print(make_spo(
Expand All @@ -546,7 +549,6 @@ def scv_link(scv_sig):
# SCV_ReviewStatus = ClinicalSignificance.find('ReviewStatus')
# if SCV_ReviewStatus is not None:
# scv_review = SCV_ReviewStatus.text
SCV_Description = ClinicalSignificance.find('Description')

SCV_Citation = \
ClinicalSignificance.find('Citation/ID[@Source="PubMed"]')
Expand All @@ -565,7 +567,8 @@ def scv_link(scv_sig):
'literature-based study'))

scv_significance = scv_geno = None
if SCV_Description:
SCV_Description = ClinicalSignificance.find('Description')
if SCV_Description is not None:
scv_significance = SCV_Description.text
scv_geno = TT[scv_significance]
if scv_geno is not None:
Expand All @@ -576,14 +579,15 @@ def scv_link(scv_sig):
monarch_assoc,
'OBAN:association_has_predicate',
scv_geno))
# store association's significance to compare w/sibs
pathocalls.update((monarch_assoc, scv_geno))
# <rcv_variant_id><scv_geno><rcv_disease_db:rcv_disease_id>
print(make_spo(rcv_variant_id, scv_geno, rcv_disease_curi))
print(make_spo('ClinVarVariant:' + rcv_variant_id, scv_geno, rcv_disease_curi))
# <monarch_assoc><OIO:hasdbxref><ClinVar:rcv_acc> .
print(make_spo(
monarch_assoc, 'OIO:hasdbxref', 'ClinVar:' + rcv_acc))

# store association's significance to compare w/sibs
pathocalls[monarch_assoc] = scv_geno

# scv_assert_type = SCV_Assertion.find('Assertion').get('Type')
# check scv_assert_type == 'variation to disease'?

Expand Down Expand Up @@ -615,7 +619,9 @@ def scv_link(scv_sig):
# blank node
_provenance_id = \
'_:' + \
hashlib.md5((_evidence_id + scv_evidence_type).encode('utf-8')).hexdigest()[1:17]
hashlib.md5(
(_evidence_id + scv_evidence_type).
encode('utf-8')).hexdigest()[1:17]
# TRIPLES
# has_provenance
# <_evidence_id><SEPIO:0000011><_provenence_id>
Expand Down Expand Up @@ -688,9 +694,7 @@ def scv_link(scv_sig):

SCV_NCBI = SCV_Measure.find('XRef[@DB="Gene"]')
if SCV_NCBI is not None:
scv_ncbigene_id = '\
NCBIGene:' + SCV_NCBI.get('ID')

scv_ncbigene_id = 'NCBIGene:' + SCV_NCBI.get('ID')
# TRIPLES
# <rcv_variant_id><GENO:0000418><scv_ncbigene_id>
print(make_spo(
Expand All @@ -713,3 +717,5 @@ def scv_link(scv_sig):
# if SCV_CiteId:
# scv_citesource = SCV_CiteId.get('Source')
# scv_citeid = SCV_CiteId.text
# print("write out any scv links for ", pathocalls)
scv_link(pathocalls)

0 comments on commit 3b0fc3e

Please sign in to comment.