How to visualize narrowPeak file after peak calling using MACS?

Question

[mudithekanayake]

Hello,

I have done peak calling for H3K27ac mark using MACS. Now I have narrowPeak file. Should I need to rearrange the file to be visualized using UCSC Genome browser? It was not working when I tried to do that using both UCSC Genome browser and IGV. Thank you.

This is how narrowPeak file looks like.

1 4491713 4491913 testONE_peak_1 37 . 2.86003 3.72580 1.66717 80 1 4492782 4493153 testONE_peak_2 27 . 2.22933 2.71107 0.82442 242 1 4571554 4572026 testONE_peak_3 72 . 4.16005 7.21000 4.85200 282 1 4739486 4739879 testONE_peak_4 64 . 3.63128 6.42948 4.14394 275 1 4784848 4785061 testONE_peak_5 31 . 2.24218 3.19160 1.21569 133

[ATpoint]

is not working is not a valid error message, just saying...

[gable_works]

Use -B during MACS2 call to create a bedGraph.

[colin.kern]

That will create a track of the normalized read pileups, but it won't show what peaks and peak boundaries were actually called by MACS2.

[gable_works]

Sure. Are peaks not called from normalized read pileups? When looking at both a bedGraph and narrowPeak file this is much more detailed. The details in a bedGraph output for "visualization" can be just as informative as peaks and boundaries. MACS2 also outputs a summit file as to indicate the peak at its highest point which is not always in the center. Your answer is correct below I'm just offering an alternative way to visualize with UCSC.

[colin.kern]

I agree with you, but the question wasn't "How should I visualize my ChIP-seq data" it was "How do I visualize a narrowPeak file" so I was just making it clear that the bedGraph file generated by -B is a different visualization showing slightly different information.

Answer 1

[colin.kern]

Your chromosome names are using the Ensembl identifiers. UCSC requires their own chromsome names to be used, which look like "chr1". You can convert the Ensembl ones to UCSC just by adding "chr" to the front, which works for all the main chromosome scaffolds. The unplaced/unlocalized ones aren't as easy, but depending on your species and your analysis, might not be important. I think the mitochondrial genome is named MT in Ensembl and chrM in UCSC. You might also need to put a UCSC track header on the file indicating the type is narrowPeak, e.g.:

track type=narrowPeak name="Track name" description="Track description"

[mudithekanayake]

Oh. Thank you very much. I'll try that. Changing the chromosome name and adding a track header will do the work right?

[colin.kern]

Maybe. You didn't provide much detail about how exactly it isn't working, so I just had to guess what your issue might be. Are you getting an error message? If the problem is what I am guessing in my answer, you will probably not be getting any errors, but the track on the browser will be empty and not show any peak regions.