def HOM_COUNT(self):
if 'HOM_COUNT' in self.__dict__: return self.__dict__['HOM_COUNT']
q=vcf.vcf_query(variant_str=self.variant_id)
if q is None: raise Exception('ERROR',self.variant_id)
self.__dict__.update(q)
print(self.save())
return self.__dict__['HOM_COUNT']
def allele_num(self):
if 'allele_num' in self.__dict__: return self.__dict__['allele_num']
q=vcf.vcf_query(variant_str=self.variant_id)
if q is None: raise Exception('ERROR',self.variant_id)
self.__dict__.update(q)
print(self.save())
return self.__dict__['allele_num']
def allele_num(self):
if 'allele_num' in self.__dict__: return self.__dict__['allele_num']
q=vcf.vcf_query(variant_str=self.variant_id)
if q is None: raise Exception('ERROR',self.variant_id)
self.__dict__.update(q)
print(self.save())
return self.__dict__['allele_num']
def HOM_COUNT(self):
if 'HOM_COUNT' in self.__dict__: return self.__dict__['HOM_COUNT']
q=vcf.vcf_query(variant_str=self.variant_id)
if q is None: raise Exception('ERROR',self.variant_id)
self.__dict__.update(q)
print(self.save())
return self.__dict__['HOM_COUNT']
def __init__(self, variant_id=None, db=None,data=None):
if variant_id is None: variant_id=data['variant_id']
self.variant_id=str(variant_id).strip().replace('_','-')
self.chrom, self.pos, self.ref, self.alt = variant_id.split('-')
q=vcf.vcf_query(variant_str=self.variant_id,)
if q is None: raise Exception('NOT IN VCF',self.variant_id)
self.__dict__.update(q)
if data: self.__dict__.update(data)
if db:
Variant.db=db
data=Variant.db.variants.find_one({'variant_id':self.variant_id},fields={'_id':False})
if not data:
print('NOT IN DB', self.variant_id, 'WILL INSERT')
self.save()
#self.xpos = get_xpos(self.chrom, self.pos)
else:
self.__dict__.update(data)
def get_rare_var_p_hpo(gene_id, db, patient_db):
#return {'hom_comp':{p_id1:{hpo: [(HP:1234,hell)], exac_af:[0.0021], uclex_af:[0.001]},
# 'het':{p_id2: {hpo:[(HP:2345,yeah)], exac_af:[0.001,0.002],uclex_af:[0.002,0.001]}}
# sometimes variant is not in vcf. move it to debug/bad_variants for inspection and later clean
bad_var_file = open('views/debug/bad_variants', 'w')
# get all variants on this gene
all_vars = db.genes.find_one({'gene_id':gene_id})['variant_ids']
results = {'hom_comp':{}, 'het':{}}
for v in all_vars:
var = db.variants.find_one({'variant_id':v})
exac_af = 0
if var['in_exac']:
if 'allele_freq' not in var['EXAC']:
VAR = annotation.exac_anno(v)
exac_af = VAR['allele_freq']
else:
exac_af = var['EXAC']['allele_freq']
# not interested if af is > 0.01
if float(exac_af) > 0.01:
continue
# get relevant info from vcf
this = vcf_query(variant_str=v)
if not this:
bad_var_file.write(v+'\n')
continue
uclex_af = this['allele_freq']
# dealing with hom patients. also add it to het. count hom as twice
# will need to deal with both_het !!! their af are different!!!
hom_p = this['hom_samples']
for p in hom_p:
populate_mode_p(results, 2, ['hom_comp', 'het'], p, exac_af, uclex_af, patient_db)
# dealing with het patients. note to check length of exac_af. longer than one?
# also added it to 'hom_comp'
het = this['het_samples']
for p in het:
results['het'][p] = results['het'].get(p, {'exac_af':[], 'uclex_af':[]})
modes = ['het']
if results['het'][p]['exac_af']:
# this patient has more than one var on this gene. copy it to hom_comp
modes.append('hom_comp')
populate_mode_p(results, 1, modes, p, exac_af, uclex_af, patient_db)
return results
def gene_hpo(gene_id):
# if gene not ensembl id then translate to
db=get_db()
db_patients=get_db('patients')
if not gene_id.startswith('ENSG'): gene_id = lookups.get_gene_by_name(get_db(), gene_id)['gene_id']
gene_name=db.genes.find_one({'gene_id':gene_id})['gene_name']
print(gene_name)
exac_thresh=request.args.get('exac_thresh')
model=request.args.get('model')
everyone=frozenset(file('/slms/UGI/vm_exports/vyp/phenotips/uclex_files/current/headers.txt','r').read().strip().split('\t'))
def condition(v):
variant_id=v['variant_id']
try:
v=Variant(db=db,variant_id=variant_id)
except:
print('NOT IN VCF',variant_id,)
return False
if v.in_exac and float(v.EXAC['allele_freq'])>0.001:
print('TOO COMMON',v.EXAC)
return False
condition=v.filter=='PASS' and v.WT_COUNT> len(everyone)/4. and ((v.HOM_COUNT==1 and v.HET_COUNT==0) or (v.HOM_COUNT<v.HET_COUNT)) and v.HET_COUNT<len(everyone)/1000.
print(v.filter)
print(v.WT_COUNT> len(everyone)/4.)
print(v.HOM_COUNT<v.HET_COUNT)
print(v.HET_COUNT<len(everyone)/1000.)
print(v.variant_id,condition)
return condition
#variants in gene
variants=[v for v in db.variants.find({'genes': gene_id}, fields={'_id': False}) if condition(v)]
#if gene_id in app.config['GENES_TO_CACHE']:
#return open(os.path.join(app.config['GENE_CACHE_DIR'], '{}.html'.format(gene_id))).read()
#else:
print(len(variants))
samples=[]
hom_samples=[]
het_samples=[]
for v in variants:
q=vcf_query(variant_str=v['variant_id'])
if not q:
print(v['variant_id'])
continue
hom_samples+=q.get('hom_samples',[])
het_samples+=q.get('het_samples',[])
hom_samples_count=Counter(hom_samples)
het_samples_count=Counter(het_samples)
print('HOM:')
print(hom_samples_count)
print('HET:')
print(het_samples_count)
if model=='recessive':
samples=frozenset([s for s in hom_samples_count]+[s for s in het_samples_count if het_samples_count[s]>1])
else:
samples=frozenset(hom_samples+het_samples)
hpo=[]
for s in everyone:
#hpo+=[f for f in db_patients.patients.find_one({'external_id':s},{'features':1}) if f['observed']=='yes']
p=db_patients.patients.find_one({'external_id':s},{'features':1})
if not p: continue
if 'features' not in p:
print(s + ' has no features ')
continue
p2=dict()
p2['features']=[f for f in p['features'] if f['observed']=='yes']
if s in samples:
p2[gene_name]=True
else:
p2[gene_name]=False
hpo.append(p2)
stats=Counter([h[gene_name] for h in hpo])
variants=[v['variant_id'] for v in variants]
return(jsonify(result=hpo,stats=stats,variants=variants))
def het_samples(self):
if 'het_samples' in self.__dict__: return self.__dict__['het_samples']
q=vcf.vcf_query(variant_str=self.variant_id)
self.__dict__.update(q)
print(self.save())
return self.__dict__['het_samples']
def het_samples(self):
if 'het_samples' in self.__dict__: return self.__dict__['het_samples']
q=vcf.vcf_query(variant_str=self.variant_id)
self.__dict__.update(q)
print(self.save())
return self.__dict__['het_samples']