A little knowledge is always a dangerous thing. I am primarily a geneticist, not a bioinformaticist, but I do like to analyze sequence data (RNAseq and ChIPseq so far) using the tools available in Galaxy - we have a local server. Anyway, I ran into a BIG problem analyzing some very nice RNAseq data from Drosophila heads. I can't seem to get RNA Star to work if I use a fasta file from Flybase combined with a GTF file from Flybase (see below for details - can provide FTP sites if you would like). I can get this to work using Ensamble's GFF3+All Fasta file, but the output data is not right - only about 40 significantly different transcript and I know that isn't true from both qRT-PCR and proteomics data we have in hand on these genotypes.
So, why do I get such variation in mapping? Why can't I run all of these combinations? Why is the GTF file not working with the Fasta file from the same FTP site and same release of the Flybase D. melanogaster genome? Any help appreciated. This may be a Galaxy specific problem, in which case I will need to get help with this analysis from our Bioinformatics core - I do not run scripts myself.
https://www.dropbox.com/s/hda2n5m2ddeqrsm/RNAStarOutput.png?dl=0
Cheers,
LTR in Memphis
