unnecessary tarball step

Hi Sukanya,

Currently tarball seems a redundant step for us. All our data have ‘fastq.gz’ format files from our sequencing provider (there are also 'fasta.gz' and '.bam' files along with it). I think directly read from those files should be enough. What I understand now is that I have to tarball all the files first and put in /data as input.

I haven't got time to look into the detail of the whole pipe, so please correct me if I'm wrong.

Cheers, Chen