vawk
vawk copied to clipboard
cc2qe
VCF header missing = silent failure
If the VCF header is missing from the input data stream, vawk seems to silently fail when constructing an awk program.
Can you show me the command you're running? I get an error with the following:
$ curl -s ftp://platgene:[email protected]/IlluminaPlatinumGenomes_v7.0/merged_platinum/NA12878.vcf.gz | gzip -cdfq | grep -v "^#" | vawk '{ print S$NA12878 }'
awk: illegal field $(), name "NA12878"
input record number 1
source line number 1
curl -s ftp://platgene:[email protected]/Illu
inaPlatinumGenomes_v7.0/merged_platinum/NA12878.vcf.gz | gzip -cdfq | grep -v "
#" | vawk '{ print S$NA12878 }' | head
chr1 116549 . C T . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 120458 . T C . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 125271 . C T . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 126113 . C A . PASS MTD=isaac2,bwa_freebaye
s,bwa_platypus,bwa_gatk3 GT 1|1
chr1 128798 . C T . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 129963 . T A . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 139967 . T C . SuspiciousHomAlt MTD=bwa
_platypus GT 1|1
chr1 172595 . G A . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 173173 . A G . SuspiciousHomAlt MTD=bwa
_freebayes GT 1|1
chr1 229673 . A C . SuspiciousHomAlt MTD=bwa
_freebayes,bwa_platypus,bwa_gatk3 GT 1|1
Another example that replicates my issue:
curl -s ftp://platgene:[email protected]/IlluminaPlatinumGenomes_v7.0/merged_platinum/NA12878.vcf.gz | gzip -cdfq | grep -v "^#" | vawk '{ if (S$NA12878$GT==S$NA12879$GT) print }' | head -n2
chr1 116549 . C T . SuspiciousHomAlt MTD=bwa_freebayes GT 1|1
chr1 120458 . T C . SuspiciousHomAlt MTD=bwa_freebayes GT 1|1
Shouldn't this fail?
And FYI, example1 works under linux, fails on my Mac. Example2 works on both linux and Mac.
Thanks for reporting these. It seems to result from awk versions behaving differently when they attempt to retrieve an key that is not in the dictionary. I'll put it on my todo list