How to use IGV (or any genome vis tool) show normalised ChIP-Seq peak intensity?
1
0
Entering edit mode
19 months ago
Tian ▴ 50

Hi everyone:

I have a question about IGV, I am quite new on ChIP-seq data.

## Question:

• In analysis side, I use MACS2 for peak calling, MAnorm2 for normalisation. etc, the result looks good.
• In IGV side, I generated bigwig file from bam files from bamCompare, then loaded these bigwig files for visaulisation.

If my understanding is correct, what IGV shows are not-normalised peak intensity.

## Two Examples

Below are two genes I identified that should have higher peak intensity value in LT and NC phenotype, and low peak intensity in the middle lane (TC). However, the IGV plot does not seems in accord with my result.

The column XX.MeanValue represent the mean normalised peak average value on the gene body. I understand the plot in IGV is using different measurements (bigwig) as analysed peak intensity (normalised log2(count)), but I am expecting they show a similar trend.


|                              | AllCount| NC.MeanValue| TC.MeanValue| LT.MeanValue|
|:-----------------------------|--------:|------------:|------------:|------------:|
...
|chr1-BTNL10                   |       24|    1.9539123|    1.1080244|    1.6104720|
|chr10-VIM                     |      153|    0.8124961|    0.5232620|    0.8462570|
...


BTNL10:

VIM:

## Solution

So my feeling is, I should try to converted normalised peak intensity into bed files, then use IGV to read these bed files, in this case, I should be able to ask IGV to show normalised peak intensity, as I right? Have anyone tried this?

IGV ChIP-seq • 1.6k views
0
Entering edit mode
19 months ago
Tian ▴ 50

I sort it out.

By loading normalised bed converted bw file, my peaks in IGV looks as good as the analysis data shows.

Below two images are the same genes as I posed like above two plots. But the tracks (bigwig) files are generated from MAnorm2 normaliased bed file.

BTNL10:

VIM:

So I know that in the future I should always use normalised data for visualisation. So that the plot is in accord with my analysis result.

I don't know why my plot looks so ugly, not smooth like many IGV plot shows...if anyone can tell me how to make smooth area plot like this one, instead of my bar-plot, I would be very much appreciate.