def get_genetic_test(cols, gene):
idx = CanRiskPedigree.get_column_idx(gene)
if idx < 0:
raise PedigreeError("Genetic test column for '" + gene +
"not found.")
gt = cols[idx].split(':')
return GeneticTest(gt[0], gt[1])
def get_genetic_test(cols, gene):
idx = CanRiskPedigree.get_column_idx(gene, file_type)
if idx < 0:
if gene == "BARD1" and file_type == "canrisk1":
return GeneticTest()
raise PedigreeError("Genetic test column for '" + gene +
"not found.")
gt = cols[idx].split(':')
return GeneticTest(gt[0], gt[1])
def validate(self):
""" Validation check for pedigree input.
@param p: Person to validate pedigree data.
"""
if (len(self.famid) > settings.MAX_LENGTH_PEDIGREE_NUMBER_STR
or not REGEX_ALPHANUM_HYPHENS.match(self.famid)
or # must be alphanumeric plus hyphen
REGEX_ONLY_HYPHENS.match(self.famid) or # but not just hyphens
REGEX_ONLY_ZEROS.match(self.famid)): # and not just zeros
raise PedigreeError(
"Family ID (1st data column) has been set to '" + self.famid +
"'. Family IDs must be specified with between 1 and " +
str(settings.MAX_LENGTH_PEDIGREE_NUMBER_STR) +
" non-zero number or alphanumeric characters.")
unconnected = self.unconnected()
if len(unconnected) > 0:
raise PedigreeError("Pedigree (" + self.famid +
") family members are not physically " +
"connected to the target: " + str(unconnected))
# Check that the index's parameters are valid
target = self.get_target()
if target.yob == '0':
raise PedigreeError(
"The target's year of birth has been set to '" + target.yob +
"'. This person must be assigned a valid year of birth.")
if target.age == '0':
raise PedigreeError("The target's age has been set to '" +
target.age +
"'. This person must be assigned an age.")
# Check that carrier probabilities / cancer risks can be computed
carrier_probs = self.is_carrier_probs_viable(target=target)
cancer_risks = self.is_risks_calc_viable(target=target)
if (not carrier_probs and not cancer_risks):
raise PedigreeError(
"BOADICEA cannot compute mutation carrier probabilities because the target '"
+ target.pid +
"' has a positive genetic test. Also BOADICEA cannot compute breast and ovarian cancer "
"risks because the target is: (1) over " +
str(settings.MAX_AGE_FOR_RISK_CALCS) +
" years old or (2) male, or (3) an affected female who has developed contralateral "
"breast cancer, ovarian cancer or pancreatic cancer.")
#
# Check monozygotic (MZ) twin data
twin_store = self.get_twins()
# Check that MZ siblings are only specified as twins, no identical triplets etc
for t in twin_store:
twins = twin_store[t]
if len(twins) != 2:
raise PedigreeError(
"MZ twin identifier '" + str(twins[0].pid) +
"' does not appear twice in the pedigree file. "
"Only MZ twins are permitted in the pedigree, MZ triplets or quads are not allowed."
)
# Check MZ twin characters are valid
if len(t) != 1 or t not in settings.UNIQUE_TWIN_IDS:
raise PedigreeError(
"Invalid MZ twin character '" + t +
"'. MZ twins must be identified using one " +
"of the following ASCII characters: " +
str(settings.UNIQUE_TWIN_IDS) + ".")
# Check that monozygotic (MZ) twin data are consistent
if (twins[0].mothid != twins[1].mothid
or twins[0].fathid != twins[1].fathid):
raise PedigreeError(
"Monozygotic (MZ) twins identified with the character '" +
t + "' have different "
"parents. MZ twins must have the same parents.")
if (twins[0].yob != twins[1].yob):
raise PedigreeError(
"Monozygotic (MZ) twins identified with the character '" +
t + "' have different "
"years of birth. MZ twins must have the same year of birth."
)
# Check that living MZ twins have the same age at last follow up
if (twins[0].dead == '0' and twins[1].dead == '0'
and twins[0].age != twins[1].age):
raise PedigreeError(
"Monozygotic (MZ) twins identified with the character '" +
t + "' have different "
"ages. If both MZ twins are alive, they must have the same age at last follow up."
)
if twins[0].sex() != twins[1].sex():
raise PedigreeError(
"Monozygotic (MZ) twins identified with the character '" +
t + "' have a different "
"sex. MZ twins must have the same sex.")
# Check that the MZ twins have the same genetic status
if not GeneticTest.compareTestResults(twins[0], twins[1]):
raise PedigreeError(
"Monozygotic (MZ) twins have both had a genetic test, but the genetic test results "
"for these individuals are different. Under these circumstances, the genetic test "
"results must be the same.")
# Check to ensure that the maximum number of MZ twin pairs per pedigree has not been exceeded
if len(twin_store.keys()) > settings.MAX_NUMBER_MZ_TWIN_PAIRS:
raise PedigreeError(
"Maximum number of MZ twin pairs has been exceeded. Input pedigrees must have a "
"maximum of " + str(settings.MAX_NUMBER_MZ_TWIN_PAIRS) +
" MZ twin pairs.")
def __init__(self,
pedigree_records=None,
people=None,
file_type=None,
bc_risk_factor_code=None,
oc_risk_factor_code=None,
bc_prs=None,
oc_prs=None):
"""
@keyword pedigree_records: the pedigree records section of the BOADICEA import pedigree file.
@keyword people: members of the pedigree.
@keyword file_type: file type is 'bwa' or 'canrisk'.
@keyword bc_risk_factor_code: breast cancer risk factor code
@keyword oc_risk_factor_code: ovarian cancer risk factor code
@keyword bc_prs: breast cancer PRS
@keyword oc_prs: ovarian cancer PRS
"""
self.people = []
if pedigree_records is not None:
self.famid = pedigree_records[0].split()[0]
ids = []
for record in pedigree_records:
p = Person.factory(record, file_type=file_type)
if p.target != '0' and p.target != '1':
raise PedigreeError(
"A value in the Target data column has been set to '" +
p.target +
"'. Target column parameters must be set to '0' or '1'."
)
if p.is_target():
self.target = p
if p.pid in ids:
raise PedigreeError(
"Individual ID '" + p.pid +
"' appears more than once in the pedigree file.")
else:
ids.append(p.pid)
self.people.append(p)
if people is not None:
self.people.extend(people)
self.famid = self.people[0].famid
ntarget = 0
for person in self.people:
if person.is_target():
ntarget += 1
pedigree_size = len(self.people)
if ntarget != 1:
raise PedigreeError(
"Pedigree (" + self.famid +
") has either no index or more than 1 " +
"index individuals. Only one target can be specified.")
if pedigree_size > settings.MAX_PEDIGREE_SIZE or pedigree_size < settings.MIN_BASELINE_PEDIGREE_SIZE:
raise PedigreeError("Pedigree (" + self.famid +
") has unexpected number of family members " +
str(pedigree_size))
if file_type == 'canrisk':
if bc_risk_factor_code is not None:
self.bc_risk_factor_code = bc_risk_factor_code
if oc_risk_factor_code is not None:
self.oc_risk_factor_code = oc_risk_factor_code
if bc_prs is not None:
self.bc_prs = bc_prs
if oc_prs is not None:
self.oc_prs = oc_prs
def factory(ped_file_line, file_type=None):
''' Factory method for creating types of people given a record from
a BOADICEA import pedigree file .
@type ped_file_line: str
@param ped_file_line: Pedigree file line.
'''
cols = ped_file_line.split()
famid = cols[0]
name = cols[1]
pid = cols[3]
cancers = Cancers(bc1=Cancer(cols[11] if cols[11] != "0" else "-1"),
bc2=Cancer(cols[12] if cols[12] != "0" else "-1"),
oc=Cancer(cols[13] if cols[13] != "0" else "-1"),
prc=Cancer(cols[14] if cols[14] != "0" else "-1"),
pac=Cancer(cols[15] if cols[15] != "0" else "-1"))
# use column headers to get gene test type and result
if file_type == 'bwa':
gtests = BWSGeneticTests.factory([
GeneticTest(cols[BwaPedigree.get_column_idx(gene + 't')],
cols[BwaPedigree.get_column_idx(gene + 'r')])
for gene in settings.BC_MODEL['GENES']
])
pathology = PathologyTests(
er=PathologyTest(PathologyTest.ESTROGEN_RECEPTOR_TEST,
cols[27]),
pr=PathologyTest(PathologyTest.PROGESTROGEN_RECEPTOR_TEST,
cols[28]),
her2=PathologyTest(PathologyTest.HER2_TEST, cols[29]),
ck14=PathologyTest(PathologyTest.CK14_TEST, cols[30]),
ck56=PathologyTest(PathologyTest.CK56_TEST, cols[31]))
else:
genes = settings.BC_MODEL['GENES'] + settings.OC_MODEL['GENES'][2:]
def get_genetic_test(cols, gene):
idx = CanRiskPedigree.get_column_idx(gene)
if idx < 0:
raise PedigreeError("Genetic test column for '" + gene +
"not found.")
gt = cols[idx].split(':')
return GeneticTest(gt[0], gt[1])
gtests = CanRiskGeneticTests.factory(
[get_genetic_test(cols, gene) for gene in genes])
path = cols[len(CanRiskPedigree.COLUMNS) - 1].split(':')
pathology = PathologyTests(
er=PathologyTest(PathologyTest.ESTROGEN_RECEPTOR_TEST,
path[0]),
pr=PathologyTest(PathologyTest.PROGESTROGEN_RECEPTOR_TEST,
path[1]),
her2=PathologyTest(PathologyTest.HER2_TEST, path[2]),
ck14=PathologyTest(PathologyTest.CK14_TEST, path[3]),
ck56=PathologyTest(PathologyTest.CK56_TEST, path[4]))
if cols[6] == 'M':
return Male(famid,
name,
pid,
fathid=cols[4],
mothid=cols[5],
target=cols[2],
dead=cols[8],
age=cols[9],
yob=cols[10],
ashkn=cols[16],
cancers=cancers,
mztwin=cols[7],
gtests=gtests,
pathology=pathology)
elif cols[6] == 'F':
return Female(famid,
name,
pid,
fathid=cols[4],
mothid=cols[5],
target=cols[2],
dead=cols[8],
age=cols[9],
yob=cols[10],
ashkn=cols[16],
cancers=cancers,
mztwin=cols[7],
gtests=gtests,
pathology=pathology)
else:
raise PedigreeError(
"The sex of family member '" + name + "' is invalid. An " +
"individuals sex must be specified as 'M' or 'F' only.")