2
$\begingroup$

I am attempting to identify DNA sequences/regions associated with enhancers in my NGS sequencing data. How can I identify (from peak files/counts for chip-seq, atac etc.) genomics regions/peaks associated with enhancers? I know proximity to genes and possibly histone (i.e. h3k27ac) activity near known genes is possibly an indicator, but is this the only way and how is the computational process formalized/captured? What are the specific set of tools involved? i.e. Peak annotations etc.? Thank you.

As I understand it in my limited bio knowledge, enhancers are at varying distances from genes and sometimes they are multiples (super-enhancers). Are all enhancers already "known/described" (akin to genes)? What is/was the process of finding them (human manual procedure)? It's that more manual/heuristic process (without ML) I am looking for so we can consider formalizing and automating it.

Based on a very insightful answer: I think some input/clarity on the following bit would make the answer(s) even more valuable:

Possible list of steps towards this include following tools/functions:

  1. Bowtie (to align fastq to a genome, mouse, human etc.)

  2. Homer/makeTagDirectory (Sep files, associate chr/coords with tags/sequence reads)

  3. Homer/findPeaks (identify and list genomic regions/sequences with count values above a threshold)

  4. Homer/pos2bed (convert peaks to bed file, genome browser compatible form)

  5. Homer/annotatePeaks.pl (Runs on *bed file - Don't know much about this, but intuitively associate a label with peaks found in 4?)

Question: Is there a step 6? Where do we go after annotatedPeaks for example to get to enhancers?

Improve this question
$\endgroup$
0

1 Answer 1

Reset to default
2
$\begingroup$

If you have reads from chip-seq or atac you can use tools like homer or macs to identify peaks in your data. Once you have the peaks, these programs allow you to do motif discovery to annotate your peaks with e.g. what transcription factor binding sites they contain.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Thank you @Niek. Very insightful contribution, that has allowed me to add specificity to the question which I have updated with your input. $\endgroup$
    – Zebra Fish
    Commented Jul 22, 2022 at 18:11

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.