fastq file with empty sequences

[lh3]

fastq file with empty sequences

[sjackman]

@AlexBardill A workaround is to replace blank lines with a single N.

sed 's/^$/N/'

[lh3]

@AlexBardill and @lexnederbragt: see 32e7903

[sjackman]

Is this issue fixed by 771d60b?

[lexnederbragt]

I could try it out, but just realised that, when one uses 'seqtk sample' on paired read files, the pairing used to be conserved by using the same seed. Is this still the case when empty sequences are ignored?

[vsbuffalo]

Yes, this should be fixed. I fixed this issue in bioawk (lh3/bioawk#9) by pulling in the newest kseq.h. I compared SHA-1 hashes of kseq.h and both bioawk and this current seqtk have the same kseq.h.

[abardill]

I'm still having this issue with the latest version of seqtk. After removing adapters with cutadapt, 0 length entries remain, which look like

@J00125:42:HCYCYBBXX:5:1106:2727:11530 2:N:0:TCCTGAGC+AGAGGATA

+

After using trimfq, this changes to

>J00125:42:HCYCYBBXX:5:1106:2727:11530 2:N:0:TCCTGAGC+AGAGGATA

[sjackman]

Thanks, Heng.