Entering edit mode
6.7 years ago
Payal
▴
160
Hi,
I am currently working on an RNA Seq dataset, which has some 3 ' bias (I found that out by Picard RNASeq Metrics). So I was suggested to perfotm 3 prime tag counting.
I came across this paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3954844/
Can someone please explain to me how they are changing the GTF based on expression and doing the length restriction?
Thanks, Payal
How biased is your data? some 3' bias is not that uncommon in any RNAseq dataset
This is how the graph looks,https://ibb.co/c7ou2H.. Is this normal?
OK, there is indeed 3' bias but would not immediately worry about it. We are talking about mapping against transcripts, right? (de novo ones or genome derived?) How did you made the graph/plot?
as Wouter said: it this consistent over all your samples?
Yes its consistent among all samples..I aligned the fastq files to reference genome via Hisat2. I used Picard RNASeq Metrics to plot this graph.
I just got to know this is FAC sorted data and it seems FAC sorted samples show this kind of 3prime bias..so i think i have to figure out how to do 3 prime tag counting!!
I wouldn't worry too much about that bias, except if there is markedly difference in the extent of the bias between samples/conditions.