Problem with the output of Deeptools PlotProfile: a strange pattern of repetitive summits

Question

[Alireza]

Hi!

I am trying to plot DNA binding profiles of my ChIP-seq bw files using Deeptools plotProfile. I generated the matrix using the computeMatrix reference-point. I used some publicly available bed files as my regions of interest, but, some of them resulted in a strange pattern as you can see below.

I aligned my data to the mm10 reference genome with the default options of bowtie2. I generated and sorted the bam files with samtools and used Picard to remove duplicates. I then used the following commands from deeptools.

bamCoverage -b  sample1.bam --effectiveGenomeSize 2494787188 --normalizeUsing RPKM -p max --extendReads 150 -o sample1.bw

computeMatrix reference-point -S sample1.bw IP.bw -R regions.bed --referencePoint center -a 3000 -b 3000 -out sample1.tab.gz --skipZeros -p max --samplesLabel sample1 IP -bs 100

I used the peak files from the following links:

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE116601

https://www-ncbi-nlm-nih-gov.proxy1.lib.uwo.ca/geo/query/acc.cgi?acc=GSE155890

Using the computeMatrix scale-regions did not solve the problem too. What is the reason for this problem and how can I solve it?

Thanks!

[rpolicastro]

Can you link to the bed files you used? You may also want to add the commands you used to align your ChIP-seq reads, and also your deepTools commands.

[Alireza]

Thanks for your reminder! I added the links and commands.

[Friederike]

What are those BED files supposed to represent? ATAC-seq peaks?

[Alireza]

No they are ChIP-seq peaks that might bind to the same regions as mine

Answer 1

[Steven]

I am having the same issue dealing with PRO-seq data, and the heatmap doesn't support this to be a true profile. Wonder if you have solve the problem or not?