Mantain mapping reads to chloroplast and mitochondrial wheat genome?

Question

[dianadd87]

Hello guys, I am new in transcriptomic analysis. I am using tophat for mapping MACE reads (Massive analysis of 3' cDNA ends) in the wheat genome. I am analyzing salt-stressed and control samples. I have realized that some reads do map to chloroplast and mitochondrial sequences in the reference... and as far as i have understood the libraries preparation consider polyA capture. I think that I should keep this reads, but i am not really sure if its better only mapping to the nuclear genome. I want to hear more opinions, and advices and additional considerations about this issue, thank you very much for your help.

[microfuge]

Hi, I had a slightly different situation (although for DNA-seq) and found that excluding anything from the reference is a very bad idea. The reference should be as complete as possible. When reference is not complete (plastid or mitochondira is excluded for example) the mapper forces reads belonging to these areas on another similar regions elsewhere and causes problems with downstream analysis particularly snp calling.