def site_specific_network_of_kinases_and_targets(f):
header = [
'kinase symbol',
'target symbol',
'kinase refseq',
'target refseq',
'target sequence position',
'target amino acid'
]
f.write('\t'.join(header) + '\n')
for protein in tqdm(Protein.query, total=fast_count(Protein.query)):
for site in protein.sites:
for kinase in site.kinases:
data = [
kinase.name,
protein.gene.name,
kinase.protein.refseq if kinase.protein else '',
protein.refseq,
site.position,
site.residue
]
f.write('\t'.join(map(str, data)) + '\n')
def mutations_affecting_ptm_sites(f, sources):
header = [
'gene',
'refseq',
'mutation position',
'mutation alt',
'mutation summary',
'site position',
'site residue'
]
f.write('\t'.join(header) + '\n')
for source in sources:
mutation_details_model = source
for mut_details in tqdm(yield_objects(mutation_details_model.query), total=fast_count(mutation_details_model.query)):
mutation = mut_details.mutation
if mutation.is_ptm():
for site in mutation.get_affected_ptm_sites():
protein = mutation.protein
summary = mut_details.summary()
data = [
protein.gene.name,
protein.refseq,
mutation.position,
mutation.alt,
', '.join(summary) if type(summary) is list else summary,
site.position,
site.residue
]
f.write('\t'.join(map(str, data)) + '\n')
def export(self, path=None, only_primary_isoforms=False):
"""Export all mutations from this source in ActiveDriver compatible format.
Source specific data export can be implemented with export_details method,
while export_details_headers should provide names for respective headers.
"""
from tqdm import tqdm
tick = 0
if not path:
path = self.generate_export_path(only_primary_isoforms)
header = [
'gene', 'isoform', 'position', 'wt_residue', 'mut_residue'
] + self.export_details_headers()
with gzip.open(path, 'wt') as f:
f.write('\t'.join(header))
for mutation in tqdm(yield_objects(self.model.query), total=fast_count(db.session.query(self.model))):
tick += 1
m = mutation.mutation
if only_primary_isoforms and not m.protein.is_preferred_isoform:
continue
dataset_specific = self.export_details(mutation)
try:
ref = m.ref
except IndexError:
print(
'Mutation: %s %s %s is exceeding the proteins sequence'
% (m.protein.refseq, m.position, m.alt)
)
ref = ''
for instance in dataset_specific:
data = [
m.protein.gene.name, m.protein.refseq,
str(m.position), ref, m.alt
] + instance
f.write('\n' + '\t'.join(data))
del data
del mutation
if tick % 10000 == 0:
import gc
gc.collect()
def interactions(self):
return (fast_count(
db.session.query(models.Site).join(
models.Kinase, models.Site.kinases)) + fast_count(
db.session.query(models.Site).join(
models.KinaseGroup, models.Site.kinase_groups)))
def kinase_groups_covered(self):
return fast_count(
db.session.query(models.KinaseGroup).filter(
models.KinaseGroup.sites.any()))
def confirmed_mutations_count(self):
return fast_count(self.confirmed_mutations)
def count(self):
return fast_count(self.query)
def export(self, path=None, only_primary_isoforms=False):
"""Export all mutations from this source in ActiveDriver compatible format.
Source specific data export can be implemented with export_details method,
while export_details_headers should provide names for respective headers.
"""
from datetime import datetime
import os
from tqdm import tqdm
export_time = datetime.utcnow()
tick = 0
if not path:
directory = os.path.join('exported', 'mutations')
os.makedirs(directory, exist_ok=True)
name_template = '{model_name}{restrictions}-{date}.tsv.gz'
name = name_template.format(
model_name=self.model_name,
restrictions=('-primary_isoforms_only'
if only_primary_isoforms else ''),
date=export_time)
path = os.path.join(directory, name)
header = ['gene', 'isoform', 'position', 'wt_residue', 'mut_residue'
] + self.export_details_headers()
with gzip.open(path, 'wt') as f:
f.write('\t'.join(header))
for mutation in tqdm(yield_objects(self.model.query),
total=fast_count(db.session.query(
self.model))):
tick += 1
m = mutation.mutation
if only_primary_isoforms and not m.protein.is_preferred_isoform:
continue
dataset_specific = self.export_details(mutation)
try:
ref = m.ref
except IndexError:
print(
'Mutation: %s %s %s is exceeding the proteins sequence'
% (m.protein.refseq, m.position, m.alt))
ref = ''
for instance in dataset_specific:
data = [
m.protein.gene.name, m.protein.refseq,
str(m.position), ref, m.alt
] + instance
f.write('\n' + '\t'.join(data))
del data
del mutation
if tick % 10000 == 0:
import gc
gc.collect()