chip seq peaks are very long

Question

[yiren]

Hello ,every one, I have a question about peaks length.I get the peaks by macs2 for chipseq data(BAMPE file),but my peaks length is very long.they almost are bigger than 500 base. this is normal right?I will list some data from my peaks file below.Help me .Thank you very much!!

chr | star | end | length |  
chrR | 1888880 | 1897200 | 8320 | .
chr3 | 1030562 | 1036279 | 5717 | .
chr4 | 15450 | 21144 | 5694 | .
chr7 | 12389 | 17091 | 4702 | .
chr5 | 667728 | 671990 | 4262 | .
chr1 | 1475634 | 1479773 | 4139 | .
chr3 | 132973 | 136895 | 3922 | .
chr5 | 347409 | 351045 | 3636 | .
chr7 | 629367 | 632930 | 3563 | .
chr5 | 708703 | 712215 | 3512 | .
chr1 | 761428 | 764877 | 3449 | .
chr1 | 3146009 | 3149230 | 3221 | .
chr1 | 3105738 | 3108941 | 3203 | .
chr7 | 385703 | 388840 | 3137 | .
chr7 | 194748 | 197879 | 3131 | .
chr2 | 623495 | 626552 | 3057 | .
chr2 | 638941 | 641974 | 3033 | .
chrR | 1722466 | 1725316 | 2850 | .
chr2 | 1649317 | 1652141 | 2824 | .
chr1 | 1952043 | 1954826 | 2783 | .
chr2 | 804016 | 806797 | 2781 | .
chr4 | 1271810 | 1274566 | 2756 | .
chr1 | 661633 | 664373 | 2740 | .
chr6 | 86747 | 89481 | 2734 | .
chr1 | 2456965 | 2459657 | 2692 | .
chr2 | 522903 | 525579 | 2676 | .
chr1 | 3045426 | 3048054 | 2628 | .
chr1 | 190996 | 193613 | 2617 | .
chrR | 1586201 | 1588812 | 2611 | .
chr4 | 1041711 | 1044314 | 2603 | .
chr6 | 76212 | 78804 | 2592 | .
chr2 | 1746703 | 1749288 | 2585 | .
chr1 | 2432263 | 2434826 | 2563 | .
chr1 | 949260 | 951816 | 2556 | .
chr7 | 9592 | 12094 | 2502 | .
chr6 | 512863 | 515365 | 2502 | .
chr1 | 1509268 | 1511753 | 2485 | .
chrR | 988412 | 990843 | 2431 | .
chr7 | 439503 | 441929 | 2426 | .
chr2 | 180741 | 183107 | 2366 | .
chr1 | 1826541 | 1828905 | 2364 | .
chr4 | 1467570 | 1469922 | 2352 | .
chrR | 1762374 | 1764697 | 2323 | .
chr1 | 1526805 | 1529087 | 2282 | .

Answer 1

[Devon Ryan]

Whether this is correct or not will depend on what you're IPing. How long do you expect it to be? Have you looked at some of the peaks and do they look reasonable? Use IGV.

[ATpoint]

Most importantly, what did you ChIP? Broad histone mark, Pol-II?

[yiren]

It is Transcription factor.

[ATpoint]

Can you show a browser track screenshot? It is unusual for TF peaks being that long. Can you also post a histogram or boxplot of the width across all peaks? Maybe you are picking up clusters of TFs in the example you show. How long were fragments after sonication? You can also check the insert size lengths in the paired-end bam file.

[yiren]

thank you for your reply.

Answer 2

[German.M.Demidov]

May be your chip-seq mark is producing wide peaks? What is the difference between narrow peak, Broad peak and a gapped peak of a chip-seq experiment and is it okay to use all of these regions for downstream analysis Or you run MACS2 with the corresponding flag. I'd recommend to check https://bioinformatics.stackexchange.com/questions/4489/determine-if-chip-seq-peaks-are-broad-or-narrow and your mark (if it is a histone mark)

[yiren]

thank you for your reply.