bwa mem for compressed fastq files

[aevgup]

Hi Everyone,

I want to use bwa mem to align my fastq files to the ref genome.
My fastq files are .bz2 compressed. Something like this a_1.fastq.bz2, a_2.fastq.bz2

Is there a way bwa mem can take compressed fastq files as input or on the fly on command line???

I am using this command

/apps/bwa mem -t 12 -v 1 -a -M /indexes/bwa/ref.fa a_1.fastq.bz2 a_2.fastq.bz2 > aln_pe.sam

This gives me an error but when I de compress it , the above command runs fine.
I have a lot of .fastq.bz2 files and would want to wrap them into a script for alignment. Is there a way I don't have to de compress the .bz2 files and still can run bwa.

Hope to hear from you soon

Regards
Varun

[lindenb]

under bash:

/apps/bwa mem -t 12 -v 1 -a -M /indexes/bwa/ref.fa \
   <(bunzip2 -c a_1.fastq.bz2) \
   <(bunzip2 -c  a_2.fastq.bz2)

[aevgup]

So basically we just have to provide input file redirection symbol?

Are parenthesis necessary to keep the command inside??

[lindenb]

yes. see http://www.tldp.org/LDP/abs/html/process-sub.html