[gringer]

I've filed a bug for this in github:

https://github.com/samtools/samtools/issues/112

In the meantime, you'll either have to use my script with the reference fasta file (in which case INDELs will change the length of the generated sequence), or add a dummy SAM line for each sequence in the reference fasta file that covers the entire range of the sequence.

[xmubingo]

I've filed a bug for this in github:

https://github.com/samtools/samtools/issues/112

In the meantime, you'll either have to use my script with the reference fasta file (in which case INDELs will change the length of the generated sequence), or add a dummy SAM line for each sequence in the reference fasta file that covers the entire range of the sequence.

Hi gringer, Thanks a lot!! I will try your code.

[xmubingo]

I've filed a bug for this in github:

https://github.com/samtools/samtools/issues/112

In the meantime, you'll either have to use my script with the reference fasta file (in which case INDELs will change the length of the generated sequence), or add a dummy SAM line for each sequence in the reference fasta file that covers the entire range of the sequence.

Hi gringer, i think find a way to solve this problem. Although the consensus sequence(cns.fa -> cns.fa) generated by vcf2fq doesn't have same length as it in reference sequence. but i find the sequence in cns.fa just lacks of some 'n' in its tail. So, we can add some 'n' to the sequences to make its length equal to its original length. for example:

cns.fa

>seq1
nnnnnnnnnnnnnnnAAAATTTTTCCCCGGGGgggccccGGTTTg

cns.fa.fixed

>seq1
nnnnnnnnnnnnnnnAAAATTTTTCCCCGGGGgggccccGGTTTgnnnnnnnnnnnnn

[jiewencai]

It's amazing

I have a slightly different processing script, but the command you have seems to work for me on my mitochondrial data (i.e. it produces a fastq file at the end of it).

I've attached my modified vcf2fq script to this post [also the fairly trivial fastq2fasta], maybe it will help to diagnose the problem.

The original vcf2fq can't producing sequence containing indels, and it's really bother me a lot,but this one can produce sequence containing indels,that's amazing.

[Yuhuan Meng]

I have a problem, if the new genome.fa contain indels, the position would be changed, if I call the genes with gtf from new genome, the called genes will not with the same regions compared with original genes in the original genome.

[gringer]

... so don't use a non-reference genome for defining the location of variants. Always link back to the reference genome when defining positions.