Entering edit mode
3.8 years ago
kanwarjag
★
1.2k
I have ScRNA-seq data for version3 prep of 10X, R2 has several Ns to the end and few seq look like as follow-
NAATTACATAAATTAGCCGTGCATCGTTGCGGGTGCCTGTAATCCCAGCTACTCAGGANNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN
NCTCTAAACAACATCCCAGAGATGGCTGGGCGCGGTGGCTCACACCTGTAATCCCAGCNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN
Will this affect mapping. Should I trim Ns but then R2 will be less than recommended 91 by 10X. Thanks for feedback.
They will affect mapping in the sense that only the real sequence is going to be mapped by aligner. Which aligner are you planning to use?
The 91 is just a recommendation that fits the expected insert size. If you want then trim it. As long as the read maps it is fine, and if it multimaps due to short size it anyway gets discarded (if using cellranger), so there is not much to worry about here. How many reads are affected. This is only troublesome if you have a notable fraction of reads with these Ns, in which case I would reach out to the facility that made the libraries.
By the way kanwarjag, you ask questions here reguarily and rarely (almost never) follow up on comments or upvote/accept answers. Please do that, it is an expected sign of appreciation towards the users who invest time to help you.
What should be the easiest way to remove all N from 3'
You can let the aligner take care of them or if you must do something then use a trimming program.
bbduk.sh
for one or any other you are familiar with. If you do that make sure you remove the corresponding reads fromR1 and I1
files to keep things in sync.