There are several popular types of RNA-Seq library prep which are frequently used: total RNA (with and without ribosomal depletion), mRNA-Seq/poly-A, and targeted mRNA-Seq/RNA exome. What would the impact of choice of these various types of RNA-Seq have on downstream fusion detection by programs like: FusionCatcher, STAR-Fusion, Arriba, and Pizzly?

For instance whist these methods have various biases and known issues, in particular rRNA depletion badly effects the accuracy of comparative exon level transcriptomics. That and RNA degradation is known to affect the efficacy of fusion detection. To date has anyone studied directly the choice of methodology on the performance of fusion detection directly, or can we infer that one method should be better than another for this usage case?

Obviously we can assume that targeted approached will only find partners of known genes, but will the different physical capture techniques introduce biases or shortfalls?