def main():
# script meant to update variants when the canonical form of a gene is changed
parser = argparse.ArgumentParser(description='Define a canonical transcript per gene when several are defined',
usage='python update_canonical_when_several.py -k md_api_key -g gene_hgnc')
parser.add_argument('-d', '--dbsnp-file', default='', required=True,
help='Path to the dbSNP file')
args = parser.parse_args()
if os.path.isfile(args.dbsnp_file):
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
curs.execute(
"SELECT a.id, b.pos, b.pos_ref, b.pos_alt, c.ncbi_name FROM variant_feature a, variant b, chromosomes c WHERE a.id = b.feature_id AND b.chr = c.name AND c.genome_version = 'hg38' AND b.genome_version = 'hg38' AND a.dbsnp_id is NULL ORDER BY a.id"
)
res = curs.fetchall()
count = curs.rowcount
log('INFO', 'Found {0} variants to check'.format(count))
i = 0
j = 0
for var in res:
j += 1
if j % 500 == 0:
log('INFO', '{0}/{1} variant checked'.format(j, count))
tb = tabix.open(args.dbsnp_file)
query = "{0}:{1}-{2}".format(var['ncbi_name'], var['pos'], var['pos'])
records = tb.querys(query)
for record in records:
match_object = re.search(r'RS=(\d+);', record[7])
if match_object:
pos_ref_list = re.split(',', record[3])
pos_alt_list = re.split(',', record[4])
if var['pos_ref'] in pos_ref_list and \
var['pos_alt'] in pos_alt_list:
new_rs_id = match_object.group(1)
# need to update var entry
curs.execute(
"UPDATE variant_feature SET dbsnp_id = %s WHERE id = %s",
(new_rs_id, var['id'])
)
log('INFO', 'Adding rsid for variant {0} - ref:{1} - alt:{2} to rs{3}'.format(var['id'], var['pos_ref'], var['pos_alt'], new_rs_id))
i += 1
db.commit()
log('INFO', '{0} rsids added'.format(i))
else:
log('ERROR', 'Your input file was not found {0}'.format(args.dbsnp_file))
def main():
parser = argparse.ArgumentParser(description='Defines NP RefSeq acc_no when lacking', usage='python update_np_acc_no.py -k ncbi_api_key')
parser.add_argument('-k', '--ncbi-api-key', default=None, required=True, help='NCBI Entrez API key.')
args = parser.parse_args()
# get file
ncbi_api_key = None
if args.ncbi_api_key is not None:
if not re.search(r'\w+', args.ncbi_api_key):
log('ERROR', 'Invalid NCBI API key, please check')
else:
ncbi_api_key = args.ncbi_api_key
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
i = 0
if ncbi_api_key is not None:
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED', ca_certs=certifi.where())
# get list of remaining genes with no canonical defined
curs.execute(
"SELECT name, np FROM gene WHERE np = 'NP_000000.0' ORDER by name"
)
res = curs.fetchall()
for acc in res:
# ncbi
ncbi_url = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=nuccore&id={0}&api_key={1}'.format(acc['name'][1], ncbi_api_key)
eutils_response = http.request('GET', ncbi_url).data.decode('utf-8')
if re.search(r'accession\s"NP_\d+",\s+version\s\d$', eutils_response, re.MULTILINE):
match_object = re.search(r'accession\s"(NP_\d+)",\s+version\s(\d+)$', eutils_response, re.MULTILINE)
curs.execute(
"UPDATE gene SET np = '{0}.{1}' WHERE name[2] = '{2}'".format(match_object.group(1), match_object.group(2), acc['name'][1])
)
log('INFO', 'Updated gene NP acc no of {0} to {1}.{2}'.format(acc['name'][0], match_object.group(1), match_object.group(2)))
i += 1
log('INFO', '{} genes updated'.format(i))
db.commit()
def main():
# script meant to update variants when the canonical form of a gene is changed
parser = argparse.ArgumentParser(
description=
'Define a canonical transcript per gene when several are defined',
usage=
'python update_canonical_when_several.py -k md_api_key -g gene_hgnc')
parser.add_argument(
'-k',
'--api-key',
default='',
required=True,
help='Your API key visible on your profile page on the website.')
parser.add_argument('-g',
'--gene-name',
default='',
required=True,
help='The gene you want to update the variants from.')
args = parser.parse_args()
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
username = None
if len(args.api_key) != 43:
log('ERROR', 'Invalid API key, please check it')
else:
api_key = args.api_key
# user
curs.execute("SELECT username FROM mobiuser WHERE api_key = %s",
(api_key, ))
res_user = curs.fetchone()
if res_user is None:
log('ERROR', 'Unknown API key')
username = res_user['username']
log('INFO', 'User: {}'.format(username))
match_obj = re.search('^([\w-]+)$', args.gene_name)
if match_obj:
gene_name = match_obj.group(1)
else:
log('ERROR', 'Invalid gene name, please check it')
# date
today = datetime.datetime.now()
creation_date = '{0}-{1}-{2}'.format(today.strftime("%Y"),
today.strftime("%m"),
today.strftime("%d"))
# check if gene exists and get new canonical isoform
curs.execute(
"SELECT DISTINCT(name[2]) as nm, nm_version FROM gene WHERE name[1] = %s AND canonical = 't'",
(gene_name, ))
res = curs.fetchone()
if res is None:
log(
'ERROR',
'The gene {} is not present in MobiDetails, please check it'.
format(gene_name))
nm = res['nm']
nm_full = '{0}.{1}'.format(res['nm'], res['nm_version'])
# get all variants
curs.execute(
"SELECT a.chr, a.pos, a.pos_ref, a.pos_alt, a.g_name, b.c_name, b.id FROM variant a, variant_feature b WHERE a.feature_id = b.id AND b.gene_name[1] = %s AND b.gene_name[2] != %s AND a.genome_version = 'hg38' ORDER BY a.pos",
(gene_name, nm))
res = curs.fetchall()
if res is None:
log('ERROR', 'No variant to update')
for var in res:
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED',
ca_certs=certifi.where())
vv_url_base = "https://rest.variantvalidator.org"
# vv_url_base = "http://0.0.0.0:8000/"
vv_url = "{0}/VariantValidator/variantvalidator/GRCh38/{1}-{2}-{3}-{4}/all?content-type=application/json".format(
vv_url_base, var['chr'], var['pos'], var['pos_ref'],
var['pos_alt'])
log('DEBUG', 'Calling VariantValidator API: {}'.format(vv_url))
try:
vv_data = json.loads(
http.request('GET', vv_url).data.decode('utf-8'))
# log('DEBUG', vv_data)
except Exception:
log('WARNING',
'No VV result for {0}:{1}'.format(nm_full, var['c_name']))
continue
for first_level_key in vv_data:
match_obj = re.search('{}:c\.(.+)$'.format(nm_full),
first_level_key)
if match_obj:
new_c_name = match_obj.group(1)
log(
'DEBUG', 'Old c_name: {0} - New c_name: {1}'.format(
var['c_name'], new_c_name))
if new_c_name == var['c_name']:
curs.execute(
"UPDATE variant_feature SET gene_name[2] = %s, \
creation_date = %s WHERE id = %s",
(nm, creation_date, var['id']))
log('INFO',
'Variant {} remains unchanged'.format(var['c_name']))
else:
# likely to change are p_name, ivs_name, prot_type, start_segment_type, start_segment_number, end_segment_type, end_segment_number
# also need to update creation_date, creation_user
# get p_name
p_name = None
if 'hgvs_predicted_protein_consequence' in vv_data[
first_level_key]:
# log('DEBUG', vv_data[first_level_key]['hgvs_predicted_protein_consequence'])
if 'tlr' in vv_data[first_level_key][
'hgvs_predicted_protein_consequence']:
# log('DEBUG', vv_data[first_level_key]['hgvs_predicted_protein_consequence']['tlr'])
match_object = re.search(
'NP_\d+\.\d.*:p\.\(?(.+)\)?',
vv_data[first_level_key]
['hgvs_predicted_protein_consequence']['tlr'])
if match_object:
p_name = match_object.group(1)
if re.search(r'\)$', p_name):
# remove last ')'
p_name = p_name[:-1]
else:
log('WARNING', 'No p_name in VV results')
else:
log('WARNING', 'No tlr in VV results')
else:
log(
'WARNING',
'No hgvs_predicted_protein_consequence in VV results'
)
start_segment_type = start_segment_number = end_segment_type = end_segment_number = ivs_name = None
# get segments type and number
positions = compute_start_end_pos(var['g_name'])
if positions[0] != positions[1]:
curs.execute(
"SELECT number, type FROM segment WHERE genome_version = 'hg38' AND \
gene_name[1] = %s AND gene_name[2] = %s AND %s BETWEEN SYMMETRIC segment_start \
AND segment_end AND %s BETWEEN SYMMETRIC segment_start AND segment_end",
(gene_name, nm, positions[0], positions[1]))
res_seg = curs.fetchone()
if res_seg is not None:
# start - end in same segment
start_segment_type = res_seg['type']
start_segment_number = res_seg['number']
end_segment_type = res_seg['type']
end_segment_number = res_seg['number']
else:
curs.execute(
"SELECT number, type FROM segment WHERE genome_version = 'hg38' \
AND gene_name[1] = %s AND gene_name[2] = %s AND %s \
BETWEEN SYMMETRIC segment_start AND segment_end ",
(gene_name, nm, positions[0]))
res_seg1 = curs.fetchone()
curs.execute(
"SELECT number, type FROM segment WHERE genome_version = 'hg38' \
AND gene_name[1] = %s AND gene_name[2] = %s AND %s \
BETWEEN SYMMETRIC segment_start AND segment_end ",
(gene_name, nm, positions[1]))
res_seg2 = curs.fetchone()
if res_strand['strand'] == '+':
start_segment_type = res_seg1['type']
start_segment_number = res_seg1['number']
end_segment_type = res_seg2['type']
end_segment_number = res_seg2['number']
else:
start_segment_type = res_seg2['type']
start_segment_number = res_seg2['number']
end_segment_type = res_seg1['type']
end_segment_number = res_seg1['number']
# get IVS name
if start_segment_type == 'intron':
ivs_obj = re.search(
r'^\d+([\+-]\d+)_\d+([\+-]\d+)(.+)$',
new_c_name)
if ivs_obj:
ivs_name = 'IVS{0}{1}_IVS{2}{3}{4}'.format(
start_segment_number, ivs_obj.group(1),
end_segment_number, ivs_obj.group(2),
ivs_obj.group(3))
else:
ivs_obj = re.search(
r'^\d+([\+-]\d+)_(\d+)([^\+-].+)$',
new_c_name)
if ivs_obj:
ivs_name = 'IVS{0}{1}_{2}{3}'.format(
start_segment_number, ivs_obj.group(1),
ivs_obj.group(2), ivs_obj.group(3))
else:
ivs_obj = re.search(
r'^(\d+)_\d+([\+-]\d+)(.+)$',
new_c_name)
if ivs_obj:
ivs_name = '{0}_IVS{1}{2}{3}'.format(
ivs_obj.group(1),
end_segment_number,
ivs_obj.group(2), ivs_obj.group(3))
else:
# substitutions
curs.execute(
"SELECT number, type FROM segment WHERE genome_version = 'hg38' \
AND gene_name[1] = %s AND gene_name[2] = %s AND %s \
BETWEEN SYMMETRIC segment_start AND segment_end ",
(gene_name, nm, positions[0]))
res_seg = curs.fetchone()
start_segment_type = res_seg['type']
start_segment_number = res_seg['number']
end_segment_type = res_seg['type']
end_segment_number = res_seg['number']
if start_segment_type == 'intron':
ivs_obj = re.search(r'^[\*-]?\d+([\+-]\d+)(.+)$',
new_c_name)
ivs_name = 'IVS{0}{1}{2}'.format(
start_segment_number, ivs_obj.group(1),
ivs_obj.group(2))
if p_name is None or \
start_segment_type is None or \
start_segment_number is None or \
end_segment_type is None or \
end_segment_number is None:
log('WARNING', 'A mandatory new parameter is lacking')
continue
if ivs_name is None:
curs.execute(
"UPDATE variant_feature SET gene_name[2] = %s, c_name = %s, p_name = %s, start_segment_type = %s, \
start_segment_number = %s, end_segment_type = %s, end_segment_number = %s, \
creation_date = %s WHERE id = %s",
(nm, new_c_name, p_name, start_segment_type,
start_segment_number, end_segment_type,
end_segment_number, creation_date, var['id']))
else:
curs.execute(
"UPDATE variant_feature SET gene_name[2] = %s, c_name = %s, p_name = %s, ivs_name = %s, start_segment_type = %s, \
start_segment_number = %s, end_segment_type = %s, end_segment_number = %s, \
creation_date = %s WHERE id = %s",
(nm, new_c_name, p_name, ivs_name,
start_segment_type, start_segment_number,
end_segment_type, end_segment_number,
creation_date, var['id']))
log(
'INFO', 'Variant {0} updated to {1}'.format(
var['c_name'], new_c_name))
db.commit()
def main():
parser = argparse.ArgumentParser(description='Define a canonical transcript per gene',
usage='python define_canonical.py [-r path/to/refGeneCanonical_2019_09_23.txt]')
parser.add_argument('-r', '--refgene', default='', required=True,
help='Path to the file containing the canonical refSeq IDs per gene (from UCSC)')
parser.add_argument('-k', '--ncbi-api-key', default=None, required=False,
help='NCBI Entrez API key. If not provided, 3rd method is not executed')
parser.add_argument('-u', '--update-refgene', default=None, required=False,
help='Update RefGene (canonical) for genes w/ on variants based on NCBI (requires NCBI API key)', action='store_true')
args = parser.parse_args()
# get file
if os.path.isfile(args.refgene):
refgeneFile = args.refgene
else:
sys.exit('ERROR: Invalid input path, please check your command')
ncbi_api_key = None
if args.ncbi_api_key is not None:
if not re.search(r'\w+', args.ncbi_api_key):
sys.exit('ERROR: Invalid NCBI API key, please check')
else:
ncbi_api_key = args.ncbi_api_key
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
i = 0
# lacking_nm = []
# first when only one isoform => canonical
log('INFO', "1st Query: genes w/ only one isoform - is automatically canonical")
curs.execute(
"SELECT name FROM gene WHERE canonical = 'f' AND name[1] IN (SELECT name[1] FROM gene GROUP BY name[1] HAVING COUNT (name[1]) = 1)"
)
res = curs.fetchall()
for acc in res:
curs.execute(
"UPDATE gene SET canonical = 't' WHERE name[2] = %s",
(acc['name'][1],)
)
# lacking_nm.append(acc['name'][0])
log('INFO', 'Updated gene {} (1st method)'.format(acc['name'][0]))
i += 1
db.commit()
# second check the refgene file
log('INFO', "2nd Query: get info from local refGene file")
for geneLine in open(refgeneFile).readlines():
# ENST - NM - gene
geneLineList = geneLine.rstrip().split("\t")
# print(geneLineList[2])
if geneLineList[2] != 'n/a' and geneLineList[2] != 'hg38.refGene.name2':
# "SELECT DISTINCT(name[1]) FROM gene WHERE name[1] = %s AND name[1] NOT IN (SELECT name[1] FROM gene WHERE canonical = 't') ORDER BY name", - removed -too long
curs.execute( # gene exists in MD (no main already set)
"SELECT DISTINCT(name[1]) FROM gene WHERE name[1] = %s ORDER BY name",
(geneLineList[2],)
)
mdgene = curs.fetchone()
if mdgene is not None:
# is not canonical?
curs.execute( # gene exists in MD (no main already set)
"SELECT canonical FROM gene WHERE name[1] = %s AND canonical = 't'",
(geneLineList[2],)
)
mdgene_can = curs.fetchone()
if mdgene_can is None:
# nm exists in md?
curs.execute(
"SELECT name FROM gene WHERE name[2] = %s",
(geneLineList[1],)
) # exists in table gene_annotation? get a nm
mdnm = curs.fetchone()
if mdnm is not None:
# ok => canonical
i += 1
postGene = '{"' + mdnm['name'][0] + '","' + mdnm['name'][1] + '"}'
# print("UPDATE gene SET canonical = 't' WHERE name = '{}'".format(postGene))
curs.execute(
"UPDATE gene SET canonical = 't' WHERE name = %s",
(postGene,)
)
log('INFO', 'Updated gene {} (2nd method)'.format(mdnm['name'][0]))
# else:
# lacking_nm.append(geneLineList[2])
# print(lacking_nm)
db.commit()
log('INFO', "3rd Query: get info from NCBI for genes with no canonical defined remaining")
# 3rd get info at NCBI
# API key mandatory
if ncbi_api_key is not None:
# get list of remaining genes with no canonical defined
# "SELECT name, np, canonical FROM gene WHERE name[1] NOT IN (SELECT name[1] FROM gene WHERE canonical='t') ORDER BY name" - removed, too long
curs.execute(
"SELECT name, np, canonical FROM gene ORDER BY name[1], canonical DESC"
)
res = curs.fetchall()
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED', ca_certs=certifi.where())
semaph_gene = None
for acc in res:
if semaph_gene != acc['name'][0]:
semaph_num_iso = 0
semaph_gene = acc['name'][0]
semaph_num_iso += 1
if semaph_num_iso > 1:
continue
# check if a canonical has been defined
# curs.execute(
# "SELECT name FROM gene WHERE canonical='t' AND name[2] = %s",
# (acc['name'][1],)
# )
# res_cano = curs.fetchone()
# if res_cano is None:
if acc['canonical'] == 'f' and \
semaph_num_iso == 1:
# ncbi
ncbi_url = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=nuccore&id={0}&api_key={1}'.format(acc['name'][1], ncbi_api_key)
eutils_response = http.request('GET', ncbi_url).data.decode('utf-8')
if re.search(r'"RefSeq\sSelect"', eutils_response):
curs.execute(
"UPDATE gene SET canonical = 't' WHERE name[2] = %s",
(acc['name'][1],)
)
i += 1
log('INFO', 'Updated gene {} (3rd method)'.format(acc['name'][0]))
if acc['np'] == 'NP_000000.0':
if re.search(r'accession\s"NP_\d+",\s+version\s\d$', eutils_response, re.MULTILINE):
match_object = re.search(r'accession\s"(NP_\d+)",\s+version\s(\d+)$', eutils_response, re.MULTILINE)
curs.execute(
"UPDATE gene SET np = '{0}.{1}' WHERE name[2] = '{2}'".format(match_object.group(1), match_object.group(2), acc['name'][1])
)
log('INFO', 'Updated gene NP acc no of {0} to {1}.{2}'.format(acc['name'][0], match_object.group(1), match_object.group(2)))
if args.update_refgene:
log('INFO', "Update refGene")
# get genes w/ no variants, and at least 2 isoforms to check which one should be canonical
curs.execute(
"SELECT name, canonical FROM gene WHERE (name[1] NOT IN \
(SELECT gene_name[1] FROM variant_feature)) AND \
(name[1] IN (SELECT name[1] FROM gene GROUP BY name[1] \
HAVING COUNT (name[1]) > 1)) ORDER BY name"
)
res = curs.fetchall()
for acc in res:
ncbi_url = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=nuccore&id={0}&api_key={1}'.format(
acc['name'][1],
ncbi_api_key
)
# log('DEBUG', ncbi_url)
eutils_response = http.request('GET', ncbi_url).data.decode('utf-8')
# if acc['name'][1] == 'NM_018257':
# log('DEBUG', eutils_response)
# log('DEBUG', acc['canonical'])
# log('DEBUG', re.search(r'"RefSeq\sSelect\scriteria"', eutils_response))
if re.search(r'"RefSeq\sSelect\scriteria"', eutils_response) and acc['canonical'] is False:
curs.execute(
"UPDATE gene SET canonical = 'f' WHERE name[1] = %s",
(acc['name'][0],)
)
# log('INFO', "UPDATE gene SET canonical = 'f' WHERE name[1] = '{}'".format(acc['name'][0]))
curs.execute(
"UPDATE gene SET canonical = 't' WHERE name[2] = %s",
(acc['name'][1],)
)
# log('INFO', "UPDATE gene SET canonical = 't' WHERE name[2] = '{}'".format(acc['name'][1]))
i += 1
log('INFO', 'Updated gene {} (4th method)'.format(acc['name'][0]))
log('INFO', '{} genes modified'.format(i))
db.commit()
def main():
parser = argparse.ArgumentParser(
description=
'Define a canonical transcript per gene and optionally updates various fields',
usage='python update_canonical_from_remote.py [-r remote_server_url]')
parser.add_argument('-r',
'--remote-server',
default='',
required=True,
help='base URL of the remote server')
parser.add_argument(
'-uca',
'--update-can-all',
default='',
required=False,
help='Optionally update canonical for all genes w/ no variants',
action='store_true')
parser.add_argument('-np',
'--update-np',
default='',
required=False,
help='Optionally update NP for genes',
action='store_true')
parser.add_argument('-uu',
'--update-uniprot',
default='',
required=False,
help='Optionally update UNIPROT IDs',
action='store_true')
parser.add_argument('-uc',
'--update-creation',
default='',
required=False,
help='Optionally update variant_creation tag',
action='store_true')
parser.add_argument(
'-un',
'--update-nm',
default='',
required=False,
help='Optionally update RefSeq Nm accession number tag',
action='store_true')
args = parser.parse_args()
remote_addr = args.remote_server
# args = parser.parse_args(['-np'])
print()
log('INFO', 'Working with server {}'.format(remote_addr))
#headers
header = {
'Accept':
'application/json',
'User-Agent':
'python-requests Python/{}.{}.{}'.format(sys.version_info[0],
sys.version_info[1],
sys.version_info[2]),
}
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
# get local list of genes with no canonical isoform defined
curs.execute(
"SELECT DISTINCT(name[1]) as hgnc FROM gene WHERE name[1] NOT IN (SELECT name[1] FROM gene WHERE canonical='t') ORDER BY name[1]"
)
no_can = curs.fetchall()
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED',
ca_certs=certifi.where())
i = 0
# lacking_nm = []
for gene in no_can:
req_url = '{0}/api/gene/{1}'.format(remote_addr, gene['hgnc'])
api_response = json.loads(
http.request('GET', req_url, headers=header).data.decode('utf-8'))
log('DEBUG', 'req_url:{0}'.format(req_url))
# log('DEBUG', 'api_response:{0}'.format(api_response))
for keys in api_response:
# log('DEBUG', 'key:{0} - value:{1}'.format(keys, api_response[keys]))
if isinstance(keys, dict) and \
'canonical' in api_response[keys]:
if api_response[keys]['canonical'] is True:
if re.search(r'NM_\d+\.\d+', keys):
match_obj = re.search(r'(NM_\d+)\.\d+', keys)
nm_acc = match_obj.group(1)
curs.execute(
"UPDATE gene set canonical = 't' WHERE name[2] = %s",
(nm_acc, ))
log('INFO', 'Updating {}'.format(nm_acc))
i += 1
db.commit()
log('INFO', '{} genes modified (canonical)'.format(i))
i = 0
if args.update_can_all:
# get genes with no variants and at least 2 isoforms to see if we need to update canonical
curs.execute(
"SELECT name, canonical FROM gene WHERE (name[1] NOT IN (SELECT gene_name[1] FROM variant_feature)) \
AND (name[1] IN (SELECT name[1] FROM gene GROUP BY name[1] HAVING COUNT (name[1]) > 1)) ORDER BY name"
)
res = curs.fetchall()
for acc in res:
req_url = '{0}/api/gene/{1}'.format(remote_addr, acc['name'][0])
api_response = json.loads(
http.request('GET', req_url,
headers=header).data.decode('utf-8'))
for keys in api_response:
if isinstance(keys, dict) and \
'canonical' in api_response[keys]:
if api_response[keys]['canonical'] is True and acc[
'canonical'] == 0:
if re.search(r'NM_\d+\.\d+', keys):
match_obj = re.search(r'(NM_\d+)\.\d+', keys)
nm_acc = match_obj.group(1)
# double check
if nm_acc == acc['name'][1]:
curs.execute(
"UPDATE gene SET canonical = 'f' WHERE name[1] = %s",
(acc['name'][0], ))
# log('INFO', "UPDATE gene SET canonical = 'f' WHERE name[1] = '{}'".format(acc['name'][0]))
curs.execute(
"UPDATE gene SET canonical = 't' WHERE name[2] = %s",
(acc['name'][1], ))
# log('INFO', "UPDATE gene SET canonical = 't' WHERE name[2] = '{}'".format(acc['name'][1]))
i += 1
log('INFO',
'Updated gene {}'.format(acc['name'][0]))
db.commit()
log('INFO', '{} genes modified (canonical all)'.format(i))
if args.update_np:
curs.execute(
"SELECT DISTINCT(name[1]) as hgnc FROM gene WHERE np = 'NP_000000.0'"
)
no_np = curs.fetchall()
j = 0
for gene in no_np:
req_url = '{0}/api/gene/{1}'.format(remote_addr, gene['hgnc'])
api_response = json.loads(
http.request('GET', req_url,
headers=header).data.decode('utf-8'))
for keys in api_response:
if isinstance(keys, dict):
if 'RefProtein' in api_response[keys] and \
api_response[keys]['RefProtein'] != 'NP_000000.0':
if re.search(r'NP_\d+\.\d+',
api_response[keys]['RefProtein']):
match_obj = re.search(r'(NM_\d+)\.\d+', keys)
nm_acc = match_obj.group(1)
np_acc = api_response[keys]['RefProtein']
curs.execute(
"UPDATE gene set np = %s WHERE name[2] = %s",
(np_acc, nm_acc))
log(
'INFO',
'Updating gene NP acc no of {0} to {1}'.format(
nm_acc, np_acc))
j += 1
db.commit()
log('INFO', '{} NP acc no modified'.format(j))
if args.update_uniprot or args.update_creation or args.update_nm:
curs.execute(
"SELECT name[1] as HGNC, name[2] as nm, nm_version, np, uniprot_id, variant_creation FROM gene ORDER BY name"
)
res = curs.fetchall()
k = l = m = n = 0
o = curs.rowcount
for gene in res:
req_url = '{0}/api/gene/{1}'.format(remote_addr, gene['hgnc'])
api_response = json.loads(
http.request('GET', req_url,
headers=header).data.decode('utf-8'))
l += 1
if l % 1000 == 0:
log('INFO', '{0}/{1} isoforms checked'.format(l, o))
for keys in api_response:
match_obj = re.search(r'^(NM_\d+)\.(\d+)$', keys)
if match_obj:
nm_acc = match_obj.group(1)
# check again
if nm_acc == gene['nm']:
if args.update_nm:
nm_version = match_obj.group(2)
# log('DEBUG', '{0}dev:{1}-prod:{2}'.format(gene['hgnc'], int(nm_version), int(gene['nm_version'])))
if int(nm_version) != int(gene['nm_version']):
# no downgrade? y => downgrade
curs.execute(
"UPDATE gene set nm_version = %s WHERE name[2] = %s",
(nm_version, nm_acc))
log(
'INFO',
'Updating gene RefSeq NM accession version of {0} from {1} to {2}'
.format(nm_acc, gene['nm_version'],
nm_version))
n += 1
if 'UNIPROT' in api_response[
keys] and args.update_uniprot:
uniprot = api_response[keys]['UNIPROT']
if uniprot != gene['uniprot_id']:
curs.execute(
"UPDATE gene set uniprot_id = %s WHERE name[2] = %s",
(uniprot, nm_acc))
log(
'INFO',
'Updating gene UNIPROT id of {0} to {1}'.
format(nm_acc, uniprot))
k += 1
if 'variantCreationTag' in api_response[
keys] and args.update_creation:
tag = api_response[keys]['variantCreationTag']
if tag != gene['variant_creation']:
curs.execute(
"UPDATE gene set variant_creation = %s WHERE name[2] = %s",
(tag, nm_acc))
log(
'INFO',
'Updating gene variantCreationTag of {0} to {1}'
.format(nm_acc, tag))
m += 1
db.commit()
log('INFO', '{} UNIPROT IDs modified'.format(k))
log('INFO', '{} variantCreationTag modified'.format(m))
log('INFO', '{} RefSeq NM accession version modified'.format(n))
def main():
parser = argparse.ArgumentParser(
description='Checks that genes accept variant creation',
usage='python check_variant_creation.py [-r remote_server_url]')
parser.add_argument('-r',
'--remote-server',
default='',
required=True,
help='base URL of the remote server')
parser.add_argument(
'-k',
'--api-key',
default='',
required=True,
help='Your API key visible on your profile page on the website.')
args = parser.parse_args()
remote_addr = args.remote_server
if re.search(r'mobidetails\.iurc', remote_addr):
log('ERROR',
'This script is not intended to work with the production server')
if len(args.api_key) != 43:
log('ERROR', 'Invalid API key, please check it')
else:
api_key = args.api_key
print()
log('INFO', 'Working with server {}'.format(remote_addr))
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
curs.execute("DELETE FROM variant_feature WHERE c_name = '1A>T'")
db.commit()
# # reinitialise gene state
# curs.execute(
# "UPDATE gene SET variant_creation = 'ok'"
# )
db.commit()
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED',
ca_certs=certifi.where())
curs.execute(
"SELECT DISTINCT(name), nm_version, variant_creation FROM gene WHERE canonical = 't' ORDER BY name"
)
# AND variant_creation IN ('hg19_mapping_default', 'hg38_mapping_default')
can = curs.fetchall()
num_can = curs.rowcount
i = 0
j = 0
k = 0
# variant = 'c.1A>T'
failed_genes = []
for gene in can:
print('.', end="", flush=True)
i += 1
if i % 500 == 0:
log('INFO', '{0}/{1} genes checked'.format(i, num_can))
# variant = '{0}.{1}:c.1A>T'.format(gene['name'][1], gene['nm_version'])
# md_url = '{0}/api/variant/create/{1}/{2}'.format(remote_addr, variant, api_key)
md_url = '{0}/api/variant/create'.format(remote_addr)
variant_chgvs = '{0}.{1}:c.1A>T'.format(gene['name'][1],
gene['nm_version'])
data = {
'variant_chgvs': urllib.parse.quote(variant_chgvs),
'caller': 'cli',
'api_key': api_key
}
# reinitialise gene state before query
curs.execute(
"UPDATE gene SET variant_creation = 'ok' WHERE name[2] = %s",
(gene['name'][1], ))
db.commit()
try:
md_response = json.loads(
http.request('POST',
md_url,
headers=md_utilities.api_agent,
fields=data).data.decode('utf-8'))
# try:
# md_response = json.loads(http.request('GET', md_url, headers={'Accept': 'application/json'}).data.decode('utf-8'))
if 'mobidetails_error' in md_response:
j += 1
log(
'WARNING',
'variant creation failed for gene {0} with error {1}'.
format(gene['name'], md_response['mobidetails_error']))
new_nm_match_obj = re.search(
r'A more recent version of the selected reference sequence NM_\d+\.\d+ is available \((NM_\d+)\.(\d+)\)',
md_response['mobidetails_error'])
if new_nm_match_obj:
nm_to_check = new_nm_match_obj.group(1)
new_ver = new_nm_match_obj.group(2)
if nm_to_check == gene['name'][1]:
curs.execute(
"UPDATE gene SET nm_version = '{0}' WHERE name[2] = '{1}'"
.format(new_ver, gene['name'][1]))
# recheck
data['variant_chgvs'] = '{0}.{1}:c.1A>T'.format(
gene['name'][1], new_ver)
# md_url_2 = '{0}/api/variant/create/{1}/{2}'.format(remote_addr, variant_2, api_key)
try:
md_response_2 = json.loads(
http.request('POST',
md_url,
headers=md_utilities.api_agent,
fields=data).data.decode('utf-8'))
# md_response_2 = json.loads(http.request('GET', md_url_2, headers={'Accept': 'application/json'}).data.decode('utf-8'))
if 'mobidetails_id' in md_response_2 and gene[
'variant_creation'] != 'ok':
curs.execute(
"UPDATE gene SET variant_creation = 'ok' WHERE name[2] = '{}'"
.format(gene['name'][1]))
continue
except Exception:
k += 1
failed_genes.append('{}'.format(gene['name'][0]))
continue
if re.search(
r'cannot be mapped directly to genome build GRCh38',
md_response['mobidetails_error']):
curs.execute(
"UPDATE gene SET variant_creation = 'hg38_mapping_default' WHERE name[2] = '{}'"
.format(gene['name'][1]))
log(
'INFO',
'MD gene table updated with variant_creation = hg38_mapping_default'
)
elif re.search(r'does not seem to map correctly to hg19',
md_response['mobidetails_error']):
curs.execute(
"UPDATE gene SET variant_creation = 'hg19_mapping_default' WHERE name[2] = '{}'"
.format(gene['name'][1]))
log(
'INFO',
'MD gene table updated with variant_creation = hg19_mapping_default'
)
elif re.search(r'with the variant position and intron',
md_response['mobidetails_error']):
curs.execute(
"UPDATE gene SET variant_creation = 'mapping_default' WHERE name[2] = '{}'"
.format(gene['name'][1]))
log(
'INFO',
'MD gene table updated with variant_creation = mapping_default'
)
elif 'mobidetails_id' in md_response and gene[
'variant_creation'] != 'ok':
curs.execute(
"UPDATE gene SET variant_creation = 'ok' WHERE name[2] = '{}'"
.format(gene['name'][1]))
db.commit()
except Exception:
log('ERROR', 'failed MD API call {}'.format(md_url))
k += 1
failed_genes.append('{}'.format(gene['name'][0]))
continue
log('INFO', '{0}/{1} genes reported a VV error'.format(j, num_can))
log('INFO', '{0}/{1} genes triggered an MD error:'.format(k, num_can))
log('INFO', failed_genes)
def main():
parser = argparse.ArgumentParser(
description='Update UNIPROT ids and protein size',
usage='python check_uniprot_ids.py [-k NCBI_API_KEY]')
parser.add_argument(
'-k',
'--ncbi-api-key',
default=None,
required=False,
help='NCBI Entrez API key. If not provided, 3rd method is not executed'
)
args = parser.parse_args()
ncbi_api_key = None
if args.ncbi_api_key is not None:
if not re.search(r'\w+', args.ncbi_api_key):
sys.exit('ERROR: Invalid NCBI API key, please check')
else:
ncbi_api_key = args.ncbi_api_key
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
i = 0
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED',
ca_certs=certifi.where())
curs.execute(
"SELECT name, np, uniprot_id, prot_size FROM gene WHERE ORDER BY name")
res = curs.fetchall()
count = curs.rowcount
i = 0
for gene in res:
# ncbi
print('.', end="", flush=True)
i += 1
if i % 500 == 0:
log('INFO', '{0}/{1} genes checked'.format(i, count))
match_obj = re.search(r'(NP_\d+)\.\d', gene['np'])
if match_obj:
# log('DEBUG', gene['name'][0])
np = match_obj.group(1)
uniprot_url = 'https://www.ebi.ac.uk/proteins/api/proteins/refseq:{}?offset=0&size=100&reviewed=true'.format(
np)
uniprot_response = json.loads(
http.request('GET',
uniprot_url,
headers={
'Accept': 'application/json'
}).data.decode('utf-8'))
# print(uniprot_response[0]['accession'])
try:
if uniprot_response[0]['accession']:
# get uniport id prot size
# print('{0}-{1}'.format(gene['uniprot_id'], uniprot_response[0]['sequence']['length']))
if gene['uniprot_id'] == uniprot_response[0]['accession']:
# print('INFO: RefSeq: {0} - {1} - {2} OK'.format(gene['np'], gene['name'][1], gene['name'][0]))
pass
else:
curs.execute(
"UPDATE gene SET uniprot_id = '{0}' WHERE name[2] = '{1}'"
.format(uniprot_response[0]['accession'],
gene['name'][1]))
# print("UPDATE gene SET uniprot_id = '{0}' WHERE name[2] = '{1}'".format(uniprot_response[0]['accession'], gene['name'][1]))
log(
'WARNING',
'Updated gene UNIPROT ID of {0} - {1} from {2} to {3}'
.format(gene['name'][0], gene['name'][1],
gene['uniprot_id'],
uniprot_response[0]['accession']))
i += 1
else:
log(
'WARNING',
'md_uniprot_id: {0} - RefSeq: {1} - {2} - {3} :not checked'
.format(gene['uniprot_id'], gene['np'],
gene['name'][1], gene['name'][0]))
except Exception:
log(
'WARNING', 'no UNIPROT ID {0} for {1} - {2}'.format(
uniprot_response, gene['name'][1], gene['name'][0]))
# get prot size from eutils
ncbi_url = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=protein&id={0}&rettype=gp&complexity=3&api_key={1}'.format(
np, ncbi_api_key)
# ncbi_url = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=nuccore&id={0}&api_key={1}'.format(gene['name'][1], ncbi_api_key)
prot_size = -1
try:
eutils_response = http.request('GET',
ncbi_url).data.decode('utf-8')
# log('DEBUG', eutils_response)
prot_match = re.search(r'Protein\s+1\.\.(\d+)',
eutils_response) #Protein\s+1\.\.(\d+)$
if prot_match:
# log('DEBUG', 'ouhou')
prot_size = prot_match.group(1)
# log('DEBUG', prot_size)
except Exception:
log(
'WARNING',
'no protein size w/ got from eutils NP acc no {0}, eutils URL:{1}'
.format(gene['np'], ncbi_url))
# log('DEBUG', prot_size)
if int(prot_size) != -1 and \
gene['prot_size'] is not None and \
int(prot_size) != int(gene['prot_size']):
curs.execute(
"UPDATE gene SET prot_size = '{0}' WHERE name[2] = '{1}'".
format(prot_size, gene['name'][1]))
log(
'WARNING',
'Updated protein size for gene {0} - {1} - {2} to {3}'.
format(gene['name'][0], gene['name'][1],
gene['uniprot_id'], prot_size))
else:
log('WARNING', 'pb w/ NP acc no {}'.format(gene['np']))
log('INFO', '{} isoforms updated'.format(i))
db.commit()
def main():
parser = argparse.ArgumentParser(
description='Insert gnomAD data into MD',
usage=
'python insert_gnomad.py [-d path/to/dir/containing/gnomad.v2.1.1.lof_metrics.by_gene.txt]'
)
parser.add_argument(
'-d',
'--directory',
default='',
required=True,
help='Path to the directory containing the gnomAD metrics by gene file'
)
args = parser.parse_args()
# get file
if os.path.isfile(args.directory):
gnomadFile = args.directory
else:
sys.exit('Invalid input path, please check your command')
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
i = 0
for geneLine in open(gnomadFile).readlines():
geneLineList = geneLine.split("\t")
# print(geneLineList[0])
curs.execute( # exists in MD?
"SELECT name FROM gene WHERE name[1] = '{0}' AND canonical = 't'".
format(geneLineList[0])
# number_of_exons IN (SELECT MAX(number_of_exons) FROM gene WHERE name[1] = '{0}')".format(geneLineList[0])
)
mdNMFirst = curs.fetchone()
if mdNMFirst is not None:
# print(mdNMFirst['nm'])
curs.execute(
"SELECT DISTINCT(gene_name[2]) FROM gene_annotation WHERE gene_name[1] = '{}'"
.format(geneLineList[0]
)) # exists in table gene_annotation? get a nm
mdNMSecond = curs.fetchone()
if mdNMSecond is None:
# does not exists => creation
i += 1
postGene = '{"' + mdNMFirst['name'][0] + '","' + mdNMFirst[
'name'][1] + '"}'
oeValues = {
'synoe': geneLineList[13],
'synlower': geneLineList[24],
'synupper': geneLineList[25],
'misoe': geneLineList[4],
'mislower': geneLineList[26],
'misupper': geneLineList[27],
'lofoe': geneLineList[23],
'loflower': geneLineList[28],
'lofupper': geneLineList[29]
}
for oeval in oeValues:
try:
oeValues[oeval] = float(oeValues[oeval])
oeValues[oeval] = "{:.2f}".format(oeValues[oeval])
except Exception:
next
curs.execute(
"INSERT INTO gene_annotation VALUES('{0}','{1}','{2}','{3}','{4}','{5}','{6}','{7}','{8}','{9}')"
.format(postGene, oeValues['synoe'], oeValues['synlower'],
oeValues['synupper'], oeValues['misoe'],
oeValues['mislower'], oeValues['misupper'],
oeValues['lofoe'], oeValues['loflower'],
oeValues['lofupper']))
log('INFO', '{} annotations added'.format(i))
db.commit()
def main():
# script meant to be croned to update NM acc versions in MD according to VariantValidator
# to be ran after uta docker update for example
# uses VV API genes2transcript
# https://rest.variantvalidator.org/VariantValidator/tools/gene2transcripts/NM_130786?content-type=application%2Fjson
vv_url_base = "https://rest.variantvalidator.org"
# vv_url_base = "http://0.0.0.0:8000/"
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
curs.execute( # get genes
"SELECT name, nm_version FROM gene WHERE canonical = 't' ORDER BY name"
)
genes = curs.fetchall()
count = curs.rowcount
i = 0
for gene in genes:
# log('DEBUG', '{}-{}'.format(gene['name'][0], i))
i += 1
if i % 500 == 0:
log('INFO', '{0}/{1} genes checked'.format(i, count))
# print("MD------{}".format(gene['name'][1]))
# get VV info for the gene
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED',
ca_certs=certifi.where())
vv_url = "{0}/VariantValidator/tools/gene2transcripts/{1}?content-type=application/json".format(
vv_url_base, gene['name'][1])
try:
vv_data = json.loads(
http.request('GET', vv_url).data.decode('utf-8'))
except Exception:
log('WARNING', 'No value for {0}'.format(gene['name'][0]))
continue
if 'transcripts' in vv_data:
# current_nm = gene['nm_version']
ts_dict = {}
for transcript in vv_data['transcripts']:
# print("VV------{}".format(transcript['reference']))
match_object = re.search(r'^(N[MR]_\d+)\.(\d{1,2})',
transcript['reference'])
if match_object:
nm_acc = match_object.group(1)
# if nm_acc == gene['name'][1]:
nm_version = match_object.group(2)
if nm_acc not in ts_dict:
ts_dict[nm_acc] = [nm_version]
else:
ts_dict[nm_acc].append(nm_version)
# do sthg with ts_dict before changing gene
for nm in ts_dict:
# exploring unconsistant NMs
curs.execute("SELECT nm_version FROM gene WHERE name[2] = %s",
(nm, ))
res_nm = curs.fetchone()
max_vv_nm = max(ts_dict[nm])
if not res_nm:
continue
# log("DEBUG", "Gene: {0} - NM: {1} - VV Max NM: {2} - MD Current NM: {3}".format(gene['name'][0], nm, max_vv_nm, res_nm[0]))
if res_nm and \
int(res_nm[0]) != int(max_vv_nm):
# NEED TO TEST IF THE TRANSCIPT WORKS!!!!!
vv_url_var = "{0}/VariantValidator/variantvalidator/GRCh38/{1}.{2}:c.1A>T/all?content-type=application/json".format(
vv_url_base, nm, max_vv_nm)
log('DEBUG',
'Calling VariantValidator API: {}'.format(vv_url_var))
try:
vv_data = json.loads(
http.request('GET',
vv_url_var).data.decode('utf-8'))
# log('DEBUG', vv_data)
except Exception:
log('WARNING',
'No VV result for {0}.{1}'.format(nm, max_vv_nm))
continue
noupdate = None
for first_level_key in vv_data:
if 'validation_warnings' in vv_data[first_level_key]:
for warning in vv_data[first_level_key][
'validation_warnings']:
if re.search(r'cannot be mapped directly to genome build', warning) or \
re.search(r'No transcript definition for', warning) or \
re.search(r'No transcripts found', warning) or \
re.search(r'expected one of', warning):
log(
'WARNING',
"Cannot update gene {0} from {1} to {2} because of {3}"
.format(gene['name'][0], res_nm[0],
max_vv_nm, warning))
noupdate = 1
break
if not noupdate:
curs.execute(
"UPDATE gene SET nm_version = %s WHERE name[2] = %s",
(max_vv_nm, nm))
log(
'INFO',
"NM UPDATE: gene {0} - {1} modified from {2} to {3}"
.format(gene['name'][0], nm, res_nm[0], max_vv_nm))
db.commit()
print('.', end="", flush=True)
def main():
parser = argparse.ArgumentParser(
description='Check isoforms differences between 2 DBs',
usage=
'python check_isoforms_differences.py -k md_api_key -nk ncbi_api_key')
parser.add_argument(
'-k',
'--api-key',
default=None,
required=True,
help='Your API key visible on your profile page on the website.')
parser.add_argument('-nk',
'--ncbi-api-key',
default=None,
required=True,
help='NCBI Entrez API key.')
parser.add_argument('-md',
'--diff-md',
default='',
required=False,
help='Check differences between MD Dev and Prod',
action='store_true')
parser.add_argument('-ncbi',
'--diff-ncbi',
default='',
required=False,
help='Check differences between NCBI RefSeq and MD',
action='store_true')
ncbi_api_key = None
args = parser.parse_args()
if args.ncbi_api_key is not None:
if not re.search(r'\w+', args.ncbi_api_key):
log('ERROR', 'Invalid NCBI API key, please check')
else:
ncbi_api_key = args.ncbi_api_key
# get db connector and cursor
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
username = None
if len(args.api_key) != 43:
log('ERROR', 'Invalid API key, please check it')
else:
api_key = args.api_key
# user
curs.execute("SELECT username FROM mobiuser WHERE api_key = %s",
(api_key, ))
res_user = curs.fetchone()
if res_user is None:
log('ERROR', 'Unknown API key')
username = res_user['username']
log('INFO', 'User: {}'.format(username))
db = get_db()
curs = db.cursor(cursor_factory=psycopg2.extras.DictCursor)
if args.diff_md:
# meant to be ran on prod server
diff = {}
# get genes w/ more than one isoform
curs.execute(
"SELECT name[1] AS hgnc, name[2] AS nm, nm_version FROM gene WHERE canonical = 't' AND name[1] IN (SELECT name[1] FROM gene GROUP BY name[1] HAVING COUNT(name[1]) > 1) ORDER by name[1]"
)
res = curs.fetchall()
i = 0
for gene in res:
i += 1
log('INFO', 'Treating gene {0} - #{1}'.format(gene['hgnc'], i))
full_nm = '{0}.{1}'.format(gene['nm'], gene['nm_version'])
base_url = "http://10.34.20.79"
md_url = '{0}/MD/api/gene/{1}'.format(base_url, gene['hgnc'])
http = urllib3.PoolManager(cert_reqs='CERT_REQUIRED',
ca_certs=certifi.where())
try:
md_data = json.loads(
http.request('GET', md_url).data.decode('utf-8'))
except Exception:
log('WARNING', 'MD not responding for {}'.format(gene['hgnc']))
if full_nm in md_data and \
md_data[full_nm]['canonical'] is True:
# log('INFO', 'No change for {}'.format(gene['hgnc']))
continue
elif full_nm not in md_data:
log('DEBUG', '{0}-{1}'.format(md_data, full_nm))
for key in md_data:
matchobj = re.search(r'^(NM_\d+)\.\d+$', key)
if matchobj:
new_nm = matchobj.group(1)
if md_data[key]['canonical'] is True:
diff[gene['hgnc']] = {
'old_can': full_nm,
'new_can': key
}
log(
'INFO',
'updating canonical for {0}: {1} instead of {2}'.
format(gene['hgnc'], key, full_nm))
curs.execute(
"UPDATE gene SET canonical = 'f' WHERE name[1] = %s",
(gene['hgnc'], ))
curs.execute(
"UPDATE gene SET canonical = 't' WHERE name[2] = %s",
(new_nm, ))
db.commit()
cmd = "python3 update_vars_when_iso_change.py -k {0} -g {1}".format(
api_key, gene['hgnc'])
returned_value = subprocess.call(cmd, shell=True)
log(
'INFO',
'Variants update returned value for {0}: {1}'.
format(gene['hgnc'], returned_value))
pp = pprint.PrettyPrinter(indent=4)
pp.pprint(diff)