Anyone using illumina2srf to generate short read archives for NCBI?

[tfseverson]

I'm trying to work out how to use illumina2srf to generate output for the NCBI short read archive.

illumina2srf seems to work well from the command line without raw files (int and nse) or *sig2.txt files required (options -R -P -c 0.6 -o). When I try to include raw files and sig2 files, illumina2srf has problems finding the files needed, although I've tried to provide the appropriate paths. I've also tried using option -I to accommodate the IPAR format for raw data, which has the txt.p.gz extensions rather than txt.gz as before the upgrade.

We have Illumina's GAII, use GAPipeline version 1.0, and rely on the IPAR to do the initial image analysis. If someone has had success with running this to prepare files for the NCBI short read archive, I would greatly appreciate an example command line.

Thanks!

[alex_kozik]

I had similar problem. It was an error message:
...
Processing tile s_1_0002_seq.txt
Couldn't load int for s_1_0002_seq.txt/0
gzip: stdout: Bad file number
Broken Pipe
...
I have examined directory with int/nse files.
I have found that int/nse files are in *.gz compressed form there.

In my case, it seems, problem was solved after uncompression of all int/nse files using gunzip prior running of illumina2srf

Let me know if it helps.

[tfseverson]

Thanks, Alex: That solved the problem.