Add Support for gzipped FASTQ Files about bowtie HOT 11 CLOSED

+1 this.
Support for input fastq as gz files would be terribly useful.
I know there are various patched versions of bowtie which add gz support, for example this one (https://github.com/earonesty/bowtie) and others, but those are all very much out of date.
I was trying to do this, I believe the necessary code would need to go into pat.h, maybe related to class PatternSource (?)... I am not at a stage yet where I could make such changes to the bowtie source - it would be great if someone could add this.

from bowtie.

This feature could be implemented in the bowtie wrapper script similar to how Bowtie 2 does it. How ever if RSEM calls the bowtie binaries directly (not recommended) we could consider implementing this at the source level. I would like to hear your thoughts on this.

from bowtie.

I do not want to hijack this issue - I am not using RSEM - I just want gz support b/c I am dealing with very large fastq files and the constant unpacking and packing takes so much time and space - shall I open a separate issue? For me an implementation in the wrapper script would be sufficient.

from bowtie.

I'm not using RSEM either, but a lot of other people do. I think implementing it at the source level is best, because it's the most general use case. Anything that makes using the software easier makes getting citations of it easier ...

from bowtie.

ch4rr0, which file in the repo is "the bowtie wrapper script" you are referring to?

from bowtie.

Update: This is coming in v1.2.1, to be released soon.

from bowtie.

is this implemented or not?

from bowtie.

Yes, Bowtie has support for compressed reads.

from bowtie.

well how then? seems like it expects a file, -c switch and reading from stdin (pipe from gunzip) doesn't seem to work.

from bowtie.

You don't need to use pipes or any unzipping programs. Just put the path to the compressed FASTQ files in the same place you would put uncompressed files. It is handled automatically by Bowtie.

Unfortunately, this information is not present in Bowtie Manual.

from bowtie.

I have a fasta file, not fastq, and it cannot index it:
$ bowtie-build genome.fa.gz genome
Input files DNA, FASTA:
genome.fa.gz
Reading reference sizes
Warning: Encountered reference sequence with only gaps
Error: No unambiguous stretches of characters in the input. Aborting...
Time reading reference sizes: 00:00:00

from bowtie.