Compatability of UCSC NCBI and ENSEMBL genomes

[Noa]

I ran I ran Cuffdiff using an ENSEMBL genome (rat RGSC3.4) and want to look at my alignments using IGV
I tried to upload the genome and gtf file I downloaded from ENSEMBL to IGV but it is not working.
If I upload my bam file to IGV using their rn4 (UCSC version of the rat genome) the alignments seem to match up with the genes. Is this kosher or will I run into problem? How compatible are the different genome versions?

Also happy to know if anyone has pros/cons of using one vs the other when doing differential expression analyses
Thanks!

[mastal]

I think the genome sequences (rn4 and RGSC3.4) should be the same, but they (UCSC/NCBI/ENSEMBL) all do their own annotations, so you may find differences in the 5' start or 3' end positions of some genes, or in some exons.

[Noa]

Thanks for the clarification. Any idea what the differences are or how different the different annotations are? Is it worth running cuffdiff with the different annotations or would it all come out the same?