def interpret_impact(args, var):
"""
Interpret the report from SnpEff or VEP to determine the impact of the variant.
SnpEff examples:
0 NON_SYNONYMOUS_CODING(MODERATE|MISSENSE|Aca/Gca|T/A|OR4F5|protein_coding|CODING|ENST00000335137|exon_1_69091_70008),
1 NON_SYNONYMOUS_CODING(MODERATE|MISSENSE|Aca/Gca|T/A|OR4F5|protein_coding|CODING|ENST00000534990|exon_1_69037_69829)
VEP examples
CSQ: Consequence|Codons|Amino_acids|Gene|hgnc|Feature|EXON|polyphen|sift
non_synonymous_codon|gaT/gaG|D/E|ENSG00000116254|CHD5|ENST00000378006|18/25|benign(0.011)|tolerated(0.3)
nc_transcript_variant|||ENSG00000116254|CHD5|ENST00000491020|5/6||
"""
impact_all = [
] # holds a list of all the transcript impacts for this variant
effect_strings_str = ""
effect_strings = []
counter = 0 # counter for anno_id
if args.anno_type == "snpEff":
try:
effect_strings_str = var.INFO["EFF"]
effect_strings = effect_strings_str.split(",")
except KeyError:
if var.INFO.has_key("SNPEFF_EFFECT"):
impact_all.append(snpEff.gatk_effect_details(var.INFO))
else:
sys.stderr.write("WARNING: The input VCF has no snpEFF annotations. " \
"Variant impact will be set to unknown\n")
for effect_string in effect_strings:
counter += 1
eff_pieces = snpEff.eff_search.findall(effect_string)
for piece in eff_pieces:
impact_string = piece[
0] # the predicted inpact, which is outside the ()
impact_detail = piece[
1] # all the other information, which is inside the ()
impact_info = snpEff.effect_map[impact_string]
impact_details = snpEff.EffectDetails(impact_string,
impact_info.priority,
impact_detail, counter,
args.maj_version)
impact_all.append(impact_details)
elif args.anno_type == "VEP":
try:
effect_strings_str = var.INFO["CSQ"]
effect_strings = effect_strings_str.split(",")
except KeyError:
sys.stderr.write("WARNING: The input VCF has no VEP annotations. \
Variant impact will be set to unknown\n")
for effect_string in effect_strings:
# nc_transcript_variant&intron_variant|||ENSG00000243485|MIR1302-11|ENST00000
each_string = effect_string.split("|")
if "&" in each_string[0]:
impact_strings = each_string[0].split("&")
# impact_strings will be [nc_transcript_variant, intron_variant]
for impact_string in impact_strings:
counter += 1
try:
impact_info = vep.effect_map[impact_string]
impact_details = vep.EffectDetails(
impact_string, impact_info.priority, effect_string,
counter)
except KeyError:
impact_details = vep.EffectDetails(
impact_string, None, effect_string, counter)
impact_all.append(impact_details)
# we expect VEP to produce a valid impact label for each_string[0]
elif "&" not in each_string[0]:
counter += 1
impact_string = each_string[0]
impact_info = vep.effect_map.get(impact_string)
try:
impact_details = vep.EffectDetails(impact_string,
impact_info.priority,
effect_string, counter)
except AttributeError:
impact_details = vep.EffectDetails(impact_string, None,
effect_string, counter)
impact_all.append(impact_details)
else:
#should not get here, as the valid -t options should be handled
#in main()
sys.exit("ERROR: Unsupported variant annotation type.\n")
return impact_all
def interpret_severe_impact(args, var):
"""
Interpret the report from SnpEff or VEP to determine the impact of the variant.
SnpEff examples:
0 NON_SYNONYMOUS_CODING(MODERATE|MISSENSE|Aca/Gca|T/A|OR4F5|protein_coding|CODING|ENST00000335137|exon_1_69091_70008),
1 NON_SYNONYMOUS_CODING(MODERATE|MISSENSE|Aca/Gca|T/A|OR4F5|protein_coding|CODING|ENST00000534990|exon_1_69037_69829)
VEP examples
CSQ: Consequence|Codons|Amino_acids|Gene|hgnc|Feature|EXON|polyphen|sift
non_synonymous_codon|gaT/gaG|D/E|ENSG00000116254|CHD5|ENST00000378006|18/25|benign(0.011)|tolerated(0.3)
nc_transcript_variant|||ENSG00000116254|CHD5|ENST00000491020|5/6||
"""
effect_strings_str = ""
effect_strings = []
impact_details = None
max_severity = 9 # initialize to a value greater than the largest value in impact info priority code
count = 0 # initialize counter for anno_id
if args.anno_type == "snpEff":
try:
effect_strings_str = var.INFO["EFF"]
effect_strings = effect_strings_str.split(",")
#print len(effect_strings)
except KeyError:
if var.INFO.has_key("SNPEFF_EFFECT"):
impact_details = snpEff.gatk_effect_details(var.INFO)
else:
sys.stderr.write("WARNING: The input VCF has no snpEFF annotations. " \
"Variant impact will be set to unknown\n")
for effect_string in effect_strings:
count += 1
eff_pieces = snpEff.eff_search.findall(effect_string)
for piece in eff_pieces:
impact_string = piece[
0] # the predicted inpact, which is outside the ()
impact_detail = piece[
1] # all the other information, which is inside the ()
impact_info = snpEff.effect_map[impact_string]
#update the impact stored only if a higher severity transcript is encountered
if impact_info.priority_code < max_severity:
impact_type = impact_info
impact_details = snpEff.EffectDetails(
impact_string, impact_info.priority, impact_detail,
count, args.maj_version)
biotype = impact_details.biotype
priority = impact_info.priority_code
max_severity = impact_info.priority_code # store the current "winning" severity for the next iteration.
else:
pass
if (biotype == "protein_coding") and (priority
== max_severity):
pass
elif (biotype != "protein_coding") and (priority
== max_severity):
impact_type = impact_info
impact_details = snpEff.EffectDetails(
impact_string, impact_info.priority, impact_detail,
count, args.maj_version)
biotype = impact_details.biotype
priority = impact_info.priority_code
max_severity = impact_info.priority_code # store the current "winning" severity for the next iteration.
elif args.anno_type == "VEP":
try:
effect_strings_str = var.INFO["CSQ"]
effect_strings = effect_strings_str.split(",")
except KeyError:
sys.stderr.write("WARNING: The input VCF has no VEP annotations. \
Variant impact will be set to unknown\n")
for effect_string in effect_strings:
# nc_transcript_variant&intron_variant|||ENSG00000243485|MIR1302-11|ENST00000
each_string = effect_string.split("|")
if "&" in each_string[0]:
impact_strings = each_string[0].split("&")
# impact_strings will be [nc_transcript_variant, intron_variant]
for impact_string in impact_strings:
count += 1
try:
impact_info = vep.effect_map[impact_string]
#update the impact stored only if a higher severity transcript is encountered
if impact_info.priority_code < max_severity:
impact_type = impact_info
impact_details = vep.EffectDetails(
impact_string, impact_info.priority,
effect_string, count)
max_severity = impact_info.priority_code # store the current "winning" severity for the next iteration.
except KeyError:
pass
# we expect VEP to produce a valid impact label for each_string[0]
elif "&" not in each_string[0]:
count += 1
impact_string = each_string[0]
impact_info = vep.effect_map.get(impact_string)
if impact_info is None:
pass
elif impact_info.priority_code < max_severity:
impact_type = impact_info
impact_details = vep.EffectDetails(impact_string,
impact_info.priority,
effect_string, count)
max_severity = impact_info.priority_code # initialize the max_severity to the former value of priority code
else:
#should not get here, as the valid -t options should be handled
#in main()
sys.exit("ERROR: Unsupported variant annotation type.\n")
return impact_details