Just running through cummerbund and it's fine, other than when I have come to pinpoint particular genes, and extract the significant ones. In the gene_exp.diff and other files prior to loading into cummerbund, the information is there, but after loading the data (cuff_data <- readCufflinks('CD-LD_out')) , the gene_id has been replaced with the other annotation. While parsing back is possible, I assume there's a way to use the original gene name in analysis.

Thanks

Annotation in CummeRbund is not the best. You can check to see if there's any annotation:

gene.features<-annotation(genes(cuff))

What I tend to do is read in the gene and isoform tracking files as tables and use columns from that to construct a mapping from XLOC ID to annotation and TCON ID to annotation.