Hi all,
I am wondering if I can pipe output from reformat.sh in BBMap into SAMtools view. My goal is to take a set of fastq and return an unaligned BAM headed with sequence dictionary (this is for use in GATK4). Using some GATK test data this part works (of course!):
bbduk.sh in1=6484_snippet_1.fastq in2=6484_snippet_2.fastq trimq=30 out=stdout.fq | reformat.sh in=stdin.fq out=stdout.bam
I suppose the question then is: can I force reformat.sh to produce a BAM without specifying ".bam" ending? Or, what form does 'stdout.bam' take and how can SAMtools recognise it? Or, can reformat.sh do the reheader itself?
All help much appreciated,
Bruce.
To answer my own Q:
bbduk.sh in1=6484_snippet_1.fastq \
in2=6484_snippet_2.fastq \
trimq=30 out=stdout.fq \
| reformat.sh in=stdin.fq \
out=stdout.bam \
| samtools view -hb - | samtools reheader 6484_snippet_header.sam - > 6484_snippet.rh.bam
Incidentally, you can skip the reformat step - bbduk.sh will allow "out=stdout.bam" also :)
Cool, I had to use addslash and underscore hence the reformat.sh, but good to know.
Ah, I see. Well, alternately, reformat.sh supports quality-trimming, so I guess it's bbduk that was unnecessary (unless you were doing a kmer-based operation).
Thought about that also, but yes, using kmers.
You have learned everything I have to teach,
My RTFM is strong (also helps that there is a good FM)
Log in to answer this question.
See Piping and Screen Output at https://github.com/BioInfoTools/BBMap/blob/master/docs/UsageGuide.txt#L18
OK, I had read that, just tried a few things and it seems you have to stream through
samtools view:returns
[W::hts_close] EOF marker is absent. The input is probably truncated.returns a reheaded BAM! Thanks for the response and making me think a bit more.