consensus sequence with samtools and bcftools 1.2

[popo]

Hi,

I want to build a consensus sequence for a small region of the genome from an individual in 1000genomes. Here's what I did:

samtools view -b ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/data/HG00096/alignment/HG00096.mapped.ILLUMINA.bwa.GBR.low_coverage.20120522.bam 17:7512445-7512545 -o HG00096_1.bam

samtools mpileup -uf ../../1kg/human_g1k_v37.fasta HG00096_1.bam | bcftools view - -o HG00096_1_w_rg.vcf

#output
[mpileup] 1 samples in 1 input files
<mpileup> Set max per-file depth to 8000

vcfutils.pl vcf2fq HG00096_1_w_rg.vcf

#error output
substr outside of string at /usr/local/bin/vcfutils.pl line 557, <> line 393.
Use of uninitialized value in lc at /usr/local/bin/vcfutils.pl line 557, <> line 393.
substr outside of string at /usr/local/bin/vcfutils.pl line 557, <> line 393.


So I had a look at the vcf file, and found that values for the alt column were always <X>, which vcfutils treated as indels. The line of code generated the errors, line 557, looked like:

substr($$seq, $beg, $end - $beg) = lc(substr($$seq, $beg, $end - $beg));

As all rows in the vcf file were treated as indels, the length of $seq didn't increase while $beg took the values of the starting position of all input lines.

Is this a bug, or it's the way vcfutils designed to behave? Is there any fix to this? It's quite possible that I was missing something very basic because I'm new to this field. Any suggestion would be helpful.

Thanks.