You can absolutely find translocations, as long as they involve the fusion of two gene products. These will be a small percentage of the total SV events in the genome. Chimerascan is one of many tools for doing this sort of thing.

You will not be able to detect copy number events, with the possible exception of very small deletions that fall entirely within an exon. Consider this: if you see more transcripts from a particular gene in your RNAseq data, is it overexpressed, copy-number amplified, or both? There's really no way to tell without corresponding evidence from the genomic DNA.