def get_hgvs_name(record, as_list=False):
"""construct the valid HGVS name as the _id field"""
chrom = record.CHROM
chromStart = record.INFO['RSPOS']
ref = record.REF
_alt_list = []
_id_list = []
_pos_list = []
for alt in record.ALT:
# should not make alt a string if alt is None
if alt:
alt = str(alt)
_alt_list.append(alt)
try:
# NOTE: current get_pos_start_end doesn't handle ALT=None case
# TODO: need to remove str(alt) when get_pos_start_end can
# handle ALT=None case
(start, end) = get_pos_start_end(chrom, chromStart, ref, alt)
_pos_list.append(OrderedDict(start=start, end=end))
# handle cases where start & end position could not be
# inferred from VCF
except ValueError:
_pos_list.append(OrderedDict(start=None, end=None))
try:
HGVS = get_hgvs_from_vcf(chrom,
chromStart,
ref,
str(alt),
mutant_type=False)
_id_list.append(HGVS)
# handle cases where hgvs id could not be inferred from vcf
except ValueError:
pass
return _id_list, _alt_list, _pos_list
def parse_one_rec(assembly, record):
"""Restructure JSON
"""
doc = {"alleles": [], "gene": [],
assembly: {},
"vartype": record.get("primary_snapshot_data").get("variant_type"),
"rsid": "rs" + str(record.get("refsnp_id")),
"dbsnp_build": int(record.get("last_update_build_id")),
"dbsnp_merges": restructure_dbsnp_merge(record.get("dbsnp1_merges")),
"citations": record.get("citations")}
data = record.get('primary_snapshot_data')
hgvs_vcf_info = get_hgvs_and_vcf(assembly,
data.get("placements_with_allele"))
allele_annotations = data.get('allele_annotations')
allele_annotations = list(allele_annotations)
doc["alleles"] = restructure_allele_freq_info(allele_annotations)
doc['gene'] = restructure_gene_info(allele_annotations)
for _item in hgvs_vcf_info:
hgvs, vcf = _item
if vcf:
doc["chrom"], pos, doc["ref"], doc["alt"] = vcf
doc["chrom"] = str(doc["chrom"])
if doc["chrom"] == "23":
doc["chrom"] = "X"
elif doc["chrom"] == "24":
doc["chrom"] = "Y"
doc[assembly] = {}
try:
if doc["vartype"] != "snv":
ref = "T" + doc["ref"]
alt = "T" + doc["alt"]
else:
ref = doc["ref"]
alt = doc["alt"]
if doc["vartype"] in ["ins", "del", "delins"]:
doc[assembly]['start'], doc[assembly]['end'] = get_pos_start_end(doc["chrom"], pos - 1, ref, alt)
else:
doc[assembly]['start'], doc[assembly]['end'] = get_pos_start_end(doc["chrom"], pos, ref, alt)
except (ValueError, AssertionError):
doc[assembly] = {}
if hgvs:
doc["_id"] = hgvs.replace('chr23', 'chrX').replace('chr24', 'chrY')
yield dict_sweep(unlist(value_convert_to_number(doc, skipped_keys=['chrom', 'ref', 'alt', 'allele', 'deleted_sequence', 'inserted_sequence'])), vals=[[], {}, None])
def get_start_end(variant_type, chrom, pos, ref, alt):
# TODO this is actually a hack.
# When the variant is not a 'snv', ref or alt might be empty.
# However `get_pos_start_end` cannot work with empty ref or alt.
# Therefore we add a preceding "T" (or any single character) to bypass this limitation.
# The detail of this hack should be handled by `get_pos_start_end` itself, not here.
if variant_type != "snv":
ref = "T" + ref
alt = "T" + alt
try:
if variant_type in ["ins", "del", "delins"]:
start, end = get_pos_start_end(chrom, pos - 1, ref, alt)
else:
start, end = get_pos_start_end(chrom, pos, ref, alt)
return start, end
except (ValueError, AssertionError):
return None, None
def annotate_start_end(hgvs_vcfs, assembly):
for hgvs_id in hgvs_vcfs:
st, end = None, None
doc = hgvs_vcfs[hgvs_id]
if 'vcf' in doc:
# remove chrom, not needed
doc['vcf'].pop('chrom', None)
try:
st, end = get_pos_start_end(
chr=None, # not even used in func
pos=doc['vcf']['position'],
ref=doc['vcf']['ref'],
alt=doc['vcf']['alt'])
if st and end:
doc[assembly] = {"start": st, "end": end}
except Exception as e:
pass
yield doc
def annotate_start_end(hgvs_vcfs, assembly):
for hgvs_id in hgvs_vcfs:
st,end = None,None
doc = hgvs_vcfs[hgvs_id]
if 'vcf' in doc:
# remove chrom, not needed
doc['vcf'].pop('chrom',None)
try:
st, end = get_pos_start_end(
chr=None, # not even used in func
pos=doc['vcf']['position'],
ref=doc['vcf']['ref'],
alt=doc['vcf']['alt'])
if st and end:
doc[assembly] = {"start": st, "end": end}
except Exception as e:
pass
yield doc
def _map_line_to_json(fields):
chrInfo = fields[0].split(":") # grch37
chrom = chrInfo[0]
chromStart = int(chrInfo[1])
ma_fin_percent = fields[7].split("/")
if fields[3]:
mutation = fields[3].split(">")
ref = mutation[0]
alt = mutation[1]
HGVS = get_hgvs_from_vcf(chrom, chromStart, ref, alt)
hg19 = get_pos_start_end(chrom, chromStart, ref, alt)
hg38 = get_pos_start_end(chrom, int(fields[30].split(":")[1]), ref, alt)
# load as json data
if HGVS is None:
return
one_snp_json = {
"_id": HGVS,
"evs":
{
"chrom": chrom,
"hg19":
{
"start": hg19[0],
"end": hg19[1]
},
"hg38":
{
"start": hg38[0],
"end": hg38[1]
},
"rsid": fields[1],
"dbsnp_version": get_dbsnp(fields[2]),
"ref": ref,
"alt": alt,
"allele_count":
{
"european_american": count_dict(fields[4]),
"african_american": count_dict(fields[5]),
"all": count_dict(fields[6])
},
"ma_fin_percent":
{
"european_american": ma_fin_percent[0],
"african_american": ma_fin_percent[1],
"all": ma_fin_percent[2]
},
"genotype_count":
{
"european_american": count_dict(fields[8]),
"african_american": count_dict(fields[9]),
"all_genotype": count_dict(fields[10])
},
"avg_sample_read": fields[11],
"gene":
{
"symbol": fields[12],
"accession": fields[13]
},
"function_gvs": fields[14],
"hgvs":
{
"coding": fields[16],
"protein": fields[15]
},
"coding_dna_size": fields[17],
"conservation":
{
"phast_cons": fields[18],
"gerp": fields[19]
},
"grantham_score": fields[20],
"polyphen2":
{
"class": polyphen(fields[21])[0],
"score": polyphen(fields[21])[1]
},
"ref_base_ncbi": fields[22],
"chimp_allele": fields[23],
"clinical_info": fields[24],
"filter_status": fields[25],
"on_illumina_human_exome_chip": fields[26],
"gwas_pubmed_info": fields[27],
"estimated_age_kyrs":
{
"ea": fields[28],
"aa": fields[29]
}
}
}
return dict_sweep(value_convert(one_snp_json), vals=["NA", "none", "unknown"])
def _map_line_to_json(fields, version):
chrInfo = fields[0].split(":") # grch37
chrom = chrInfo[0]
chromStart = int(chrInfo[1])
ma_fin_percent = fields[7].split("/")
if fields[3]:
mutation = fields[3].split(">")
ref = mutation[0]
alt = mutation[1]
hg19 = get_pos_start_end(chrom, chromStart, ref, alt)
hg38 = get_pos_start_end(chrom, int(fields[30].split(":")[1]), ref,
alt)
if version == 'hg19':
HGVS = get_hgvs_from_vcf(chrom, chromStart, ref, alt)
elif version == 'hg38':
HGVS = get_hgvs_from_vcf(chrom, hg38[0], ref, alt)
# load as json data
if HGVS is None:
return
one_snp_json = {
"_id": HGVS,
"evs": {
"chrom": chrom,
"hg19": {
"start": hg19[0],
"end": hg19[1]
},
"hg38": {
"start": hg38[0],
"end": hg38[1]
},
"rsid": fields[1],
"dbsnp_version": get_dbsnp(fields[2]),
"ref": ref,
"alt": alt,
"allele_count": {
"european_american": count_dict(fields[4]),
"african_american": count_dict(fields[5]),
"all": count_dict(fields[6])
},
"ma_fin_percent": {
"european_american": ma_fin_percent[0],
"african_american": ma_fin_percent[1],
"all": ma_fin_percent[2]
},
"genotype_count": {
"european_american": count_dict(fields[8]),
"african_american": count_dict(fields[9]),
"all_genotype": count_dict(fields[10])
},
"avg_sample_read": fields[11],
"gene": {
"symbol": fields[12],
"accession": fields[13]
},
"function_gvs": fields[14],
"hgvs": {
"coding": fields[16],
"protein": fields[15]
},
"coding_dna_size": fields[17],
"conservation": {
"phast_cons": fields[18],
"gerp": fields[19]
},
"grantham_score": fields[20],
"polyphen2": {
"class": polyphen(fields[21])[0],
"score": polyphen(fields[21])[1]
},
"ref_base_ncbi": fields[22],
"chimp_allele": fields[23],
"clinical_info": fields[24],
"filter_status": fields[25],
"on_illumina_human_exome_chip": fields[26],
"gwas_pubmed_info": fields[27],
"estimated_age_kyrs": {
"ea": fields[28],
"aa": fields[29]
}
}
}
return dict_sweep(value_convert(one_snp_json),
vals=["NA", "none", "unknown"])
