1
$\begingroup$

I would like to calculate what is the number of region which is lost or gained between two condition,such as Stem cell vs Leukemic stemcell, atac seq data.

How do I get the information after peak calling such as what tools should i be using ?

As shown in the paper figure 1c I would like to do the same. How do i calculate the gain and loss of peak relative to a condition paper

Improve this question
$\endgroup$

1 Answer 1

Reset to default
3
$\begingroup$

What you are looking for is "Differential accessibility analysis", you are using ATAC-seq, hence "accessibility". You will need to "call" and then "normalize" "count"(*) peaks, after these, you can use DESeq2 or edgeR for the "differential" part.

Here is a nice Nextflow workflow that addresses all of the steps of a differential accessibility analysis of ATAC-Seq data.

* featureCounts can be used the identify peaks, see this and this.

Improve this answer
$\endgroup$
8
  • $\begingroup$ "You will need to "call" and then "normalize" peaks" this part im done. I tried with diffbind, but i was not sure about the result output clearly. Yes will check the nexflow one which i did see but haven;t tried it yet $\endgroup$
    – kcm
    Commented Nov 19, 2019 at 12:49
  • 1
    $\begingroup$ I think I was wrong about the normalization part. DESeq2 and edgeR works on raw counts so I believe you will need to use featureCounts to get the counts to feed into one of the two DE tools. $\endgroup$
    – haci
    Commented Nov 19, 2019 at 13:04
  • $\begingroup$ "ncbi.nlm.nih.gov/pmc/articles/PMC6527694/#MOESM2" preclisely yes they need raw count and how do i get raw counts from peaks? i ran deseq2 from rna seq data using raw count from bam files , i guess i can't do the same for ATAC seq ,need to get the counts from the peaks isn;t it? $\endgroup$
    – kcm
    Commented Nov 19, 2019 at 13:07
  • 1
    $\begingroup$ You can use featureCounts, I am editing my answer so as to include two links from the Bioconductor forum. $\endgroup$
    – haci
    Commented Nov 19, 2019 at 13:15
  • $\begingroup$ I manage to get differential peak analysis using deseq2 , now back to my original question now I do have differential peak between various stages how do I define the loss and gain of peaks between stages from deseq2 results or I have to use bedtool or bedops in this case ? $\endgroup$
    – kcm
    Commented Feb 17, 2020 at 17:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.