Closed GoogleCodeExporter closed 9 years ago
If you have, say, 2 fastq's , you can do this:
rm fif; mkfifo fif;
( gunzip -c fq1.gz >> fif; gunzip -c fq1.gz >> fif ) &
fastq-mcf adap.txt fif > out
Unlike much bioinformatics software, fastq-mcf is "stream friendly" ....
meaning it buffers what it needs during initial sampling steps.
Original comment by earone...@gmail.com
on 9 Oct 2013 at 2:41
Original comment by earone...@gmail.com
on 9 Oct 2013 at 2:41
Thanks.
Thanks for this very useful tip.
It can also be done without explicitly making named pipes, but by using process
substitution:
http://tldp.org/LDP/abs/html/process-sub.html
fastq-mcf \
-o cleaned.R1.fq.gz \
-o cleaned.R2.fq.gz \
adapters.fa \
<(zcat uncleaned.lane1.R1.fq.gz uncleaned.lane2.R1.fq.gz;) \
<(zcat uncleaned.lane1.R2.fq.gz uncleaned.lane2.R2.fq.gz;)
fastq-mcf \
-o cleaned.R1.fq.gz \
-o cleaned.R2.fq.gz \
adapters.fa \
<(gunzip -c uncleaned.lane1.R1.fq.gz uncleaned.lane2.R1.fq.gz;) \
<(gunzip -c uncleaned.lane1.R2.fq.gz uncleaned.lane2.R2.fq.gz;)
Can this info being added to the wiki?
http://code.google.com/p/ea-utils/wiki/FastqMcf
Thanks
Original comment by hulselma...@gmail.com
on 10 Oct 2013 at 8:05
Cool! I didn't even know how process substitution worked.
Original comment by earone...@gmail.com
on 10 Oct 2013 at 2:24
Original issue reported on code.google.com by
hulselma...@gmail.com
on 9 Oct 2013 at 1:21