Entering edit mode
7.2 years ago
bxia
▴
180
Tophat2 very high percentage of multiple alignments
About 50%-70% reads have multiple alignments, any possible parameters to solve the problem?
Thanks
Have you checked to see if they are rRNA? If your sample was not depleted/the depletion did not work right then this observation shows up.
It's also important to ask what you are trying to do. If this is a smallRNA experiment, or your reads are 36bp single-ended, or you're mapping to the transcriptome, such a high number is expected, depending on the organism. So, can you explain your experiment and procedure?
No, it is just regular pair-end RNA-seq, it is zebrofish.
In that case, and in the absence of any further description of your experiment, I support Genomax's suggestion of rRNA reads overwhelming your reads of interest. Most organisms have multiple, highly-expressed rRNA copies that constitute an inordinate amount of their RNA volume; they are typically very similar (usually 99% or higher identity in bacteria, and somewhere around there in fungi). This leads to multi-mapping reads in RNA-seq experiments with insufficient rRNA depletion.
Thank you very much, I think I will just go ahead and analyze the data.