Trouble running Tophat & Cufflinks on Galaxy with non-UCSC-supported genome

[NSher]

I have been trying to run tophat and cufflinks on Prochlorococcus genome (marine cyanobacteria). I am working using Galaxy; I am not a Linux user but rather a "wet' biologist trying to get into bioinformatics on my own).

I used the FASTA format of the genome available on MicrobesOnline: http://www.microbesonline.org/cgi-bi....cgi?tId=59919

I used a gtf file that the microbesonline people helped me create from the tab delimited version available on their website.

However, I am getting FPKMs of zero for everything.

I wasnt sure whether there is some disconnect between the genome and transcriptome files I am using, or whether the problem is that I am using a regular FASTA genome but not a bowtie index, or something else??? I would really appreciate troubleshooting help as there is very little help out there for people working on non-UCSC organisms!

[tnabtaf]

I used the new "USeGalaxy" custom search (http://galaxyproject.org/search/usegalaxy/) and, um, didn't find anything either.

Try reposting the question to the Galaxy-User mailing list. You can subscribe here: http://lists.bx.psu.edu/listinfo/galaxy-user

[rnaseek]

Check this, it might solve the problem. Tophat expects chromosome/feature names used in the annotation file to be the same as used in genome sequence/bowtie indices (and possibly same order). For example, if the chromosomes are named like "chr1" in the indices , the annotation file should also have "chr1" (first column) not "1" or "c1". you can check how the chromosomes are named using the command
"bowtie-inspect --names path2index". And change the annotation file to be the same as in the index files.