How to solve the "tilted tail" after DESeq2 normalization.
1
0
Entering edit mode
changxu.fan ▴ 20
@changxufan-24128
Last seen 12 months ago
United States

ATAC-seq samples from 2 different tissues were collected, with 2 replicates each, and normalized using DESeq2 estimateSizeFactors. The scatter plot below shows the normalized counts for each peak, with a blue color scale indicating point density where points are too dense to see. The mean of samples from each tissue (BM and SP) on x and y axes. The blue line indicates x~y. Unfortunately, at the higher end (the "tail"), most points (peaks) are below the y~x line, causing DESeq2 to identify many differentially accessible peaks higher in BM (DAPs, shown in red). The tilted tail most likely reflects a batch effect, because samples from the 2 tissues were collected from 2 independent batches of animals. We also have previous samples where samples from the 2 tissues were collected from the same animal at the same time, and the same scatter plot did not show the tail tilting issue.

I was wondering if there is a DESeq2 built-in on 3rd party method to correct/avoid the "drifted" tail?

Best, Changxu

enter image description here

DESeq2 • 650 views
ADD COMMENT
3
Entering edit mode
@mikelove
Last seen 5 hours ago
United States

You could use the larger count peaks for defining size factors, e.g..

cg <- which(rowSums(counts(dds) > x) > y)
dds <- estimateSizeFactors(dds, controlGenes=cg)
ADD COMMENT

Login before adding your answer.

Traffic: 650 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6