filtering or not the gtf annotation file from ncRNA, tRNA, and rRNA before mapping ?

Question

[debitboro]

Hi biostars community,

I've conducted a DE analysis on RNASeq data. In the first time, I realized the study using a filtered annotation file (.gtf file downloaded from Ensembl and in which I removed all the ncRNAs, tRNAs, and rRNAs) in the mapping step. In a second time I left the annotation file without any modification. As a results from the two pipelines, I got 1500 DE features in the 1st case, and more than 7000 DE features in the second case. Which of the two pipelines gives good results ?

waiting for your suggestions

thanks

[shunyip]

I believe you need to check whether those 5500 extra DE are all nc, t or r RNAs.