def transform_ref_seq(self, to_type):
'''
Transform the reference sequence to target type
:param to_type: type of reference sequence
:return:
'''
if self.ref_type == to_type:
return self.ref_seq
''' All transformation: myvariant'''
# MyVariantInfo(url='http://myvariant.info/v1')
mv_res = self.get_myvariant_res()
if mv_res:
res = MyVariantUtil.extract(mv_res, to_type)
if res:
return res
''' Fall back : EnsemblTranscriptDict '''
if to_type == GeneVariant.REF_TYPE_ENST or self.ref_type == GeneVariant.REF_TYPE_ENST:
''' If ENST transformation, first try DB:EnsemblTranscriptDict '''
cancer_db = config.connect_db(config.CANCER_DB_KEY)
from_ENST = self.ref_type == GeneVariant.REF_TYPE_ENST
query = "SELECT Ref_Seq FROM EnsemblTranscriptDict WHERE ENST_ID = %s" \
if from_ENST else "SELECT ENST_ID FROM EnsemblTranscriptDict WHERE Ref_Seq = %s"
result_key = 'Ref_Seq' if from_ENST else 'ENST_ID'
params = (self.ref_seq, )
cursor = cancer_db.cursor()
cursor.execute(query, params)
res = cursor.fetchone()
if res:
to_ref_seq = res.get(result_key, None)
if to_ref_seq and GeneReference.get_ref_type(
to_ref_seq) == to_type:
return to_ref_seq
return None
def search_variant_db(variant, only_count=False, start=0, limit=0):
res = None
if isinstance(variant, GeneVariant):
cancer_db = config.connect_db(config.CANCER_DB_KEY)
cursor = cancer_db.cursor()
# try myvariant directly
mv_res = variant.get_myvariant_res()
if mv_res:
cosmic_id = MyVariantUtil.extract(mv_res, 'cosmic.cosmic_id')
if cosmic_id:
base_query = "SELECT " + ("COUNT(*) count" if only_count else "*") \
+ " FROM COSMICMutantExport WHERE ID_Mutation=%s"
base_params = (cosmic_id,)
query, params = util_query.add_limit(base_query, base_params, start, limit)
cursor.execute(query, params)
res = cursor.fetchall()
if not res or (only_count and res[0]['count'] == 0):
# try chromosome position search
chr_pos = variant.transform_variant(GeneVariant.TRANSFORM_CHR_POS)
mutation_genome_pos = str(map_chr_to_num(chr_pos[0])) + ':' + chr_pos[1]
base_query = "SELECT " + ("COUNT(*) count" if only_count else "*") \
+ " FROM COSMICMutantExport WHERE Mutation_Genome_Position LIKE %s"
base_params = (mutation_genome_pos + '%',)
query, params = util_query.add_limit(base_query, base_params, start, limit)
cursor.execute(query, params)
res = cursor.fetchall()
if not res or (only_count and res[0]['count'] == 0):
var_infos = [variant.transform_variant(variant.ref_type, variant.INFO_TYPE_C)[1],
map_aa_3to1(variant.transform_variant(variant.ref_type, variant.INFO_TYPE_P)[1])]
info_attrs = ['Mutation_CDS', 'Mutation_AA']
ref_seqs = [variant.transform_ref_seq(GeneVariant.REF_TYPE_ENST),
variant.transform_ref_seq(GeneVariant.REF_TYPE_NM)]
found = False
for i in range(0, 2):
if found:
break
var_info = var_infos[i]
info_attr = info_attrs[i]
base_query = "SELECT " + ("COUNT(*) count" if only_count else "*") \
+ " FROM COSMICMutantExport WHERE Accession_Number=%s AND " + info_attr + " LIKE %s"
for ref_seq in ref_seqs:
if ref_seq:
base_params = (ref_seq, var_info,)
query, params = util_query.add_limit(base_query, base_params, start, limit)
cursor.execute(query, params)
res = cursor.fetchall()
if (not only_count and res) or (only_count and res[0]['count'] > 0):
found = True
break
if res:
return res if not only_count else res[0]['count']
return [] if not only_count else 0
def search_variant_db(variant):
if isinstance(variant, GeneVariant):
chr_pos_ref_alt = variant.transform_variant(
GeneVariant.TRANSFORM_CHR_POS_REF_ALT)
if chr_pos_ref_alt:
gnomad_id = '-'.join(chr_pos_ref_alt)
chr_num = map_chr_to_str(chr_pos_ref_alt[0])
table_name = 'gnomad_variants_info_sites_' + chr_num
query = "SELECT * FROM " + table_name + " WHERE Variant_ID=%s "
params = (gnomad_id, )
cancer_db = config.connect_db(config.CANCER_DB_KEY)
cursor = cancer_db.cursor()
cursor.execute(query, params)
res = cursor.fetchone()
return res
return None
def search_gene_db(reference, only_count=False, start=0, limit=0):
res = None
if isinstance(reference, GeneReference):
gene = reference.ref_seq
if reference.ref_type != GeneVariant.REF_TYPE_GENE:
gene = reference.transform_ref_seq(GeneVariant.REF_TYPE_GENE)
if gene:
cancer_db = config.connect_db(config.CANCER_DB_KEY)
cursor = cancer_db.cursor()
base_query = "SELECT " + ("COUNT(*) count" if only_count else "*") \
+ " FROM COSMICMutantExport WHERE Gene_Name LIKE %s"
base_params = (gene,)
query, params = util_query.add_limit(base_query, base_params, start, limit)
cursor.execute(query, params)
res = cursor.fetchall()
if res:
return res if not only_count else res[0]['count']
return [] if not only_count else 0
def search_transcript_db(reference, only_count=False, start=0, limit=0):
res = None
if isinstance(reference, GeneReference):
base_query = "SELECT " + ("COUNT(*) count" if only_count else "*") \
+ " FROM COSMICMutantExport WHERE Accession_Number LIKE %s"
cancer_db = config.connect_db(config.CANCER_DB_KEY)
cursor = cancer_db.cursor()
ref_seq1 = reference.transform_ref_seq(GeneVariant.REF_TYPE_ENST)
ref_seq2 = reference.transform_ref_seq(GeneVariant.REF_TYPE_NM)
for ref_seq in (ref_seq1, ref_seq2):
if ref_seq:
base_params = (ref_seq,)
query, params = util_query.add_limit(base_query, base_params, start, limit)
cursor.execute(query, params)
res = cursor.fetchall()
if res:
break
if res:
return res if not only_count else res[0]['count']
return [] if not only_count else 0