Hi all,

This is going to be a very long message, so please bear with me. Anyway...

I have been asked by my colleagues to integrate modified versions of velveth, velvetg and oases into our Galaxy instance. Of course, the first step was to upload the would-be input files. Since these are stored as .dat files in the database/files/000 folder, I symlinked the uploads to another folder. When I run the commands of the aforementioned steps, they access these symlinked files and perform the needed operations using them. I then just copy the data from the generated files onto the output files that would be visible in Galaxy. Is everything still clear? I hope so. Moving on...

Now, they have asked me to integrate another script, and this time it is running rsem. It requires corrected left and right reads, and OASES' transcript output. But when I ran the rsem, I got an error concerning the input transcript file. I found two probable reasons why I was getting an error. One was that the transcript file might be empty. I've already checked this, and the file was not empty. Another reason was that the transcript FASTA file might not be following the required format. And so we arrive at my question:

Do input and output files go through some minor format modifications when used in Galaxy?

I have already manually ran the above commands, and everything works well. But when I downloaded the transcript file from Galaxy to use it via the konsole, I got the same error.

Thank you in advance for any help,

CL