Differential Regions In Histone Modification

Question

[ChIP]

Hi!

I have a set of .BAM files (with reads) on different histone marks but I do not have replicates.

The files that I have are H3K27ac (untreated), H3K27ac (treated with sample1) and H3K27ac (treated with sample2).

I have already called peaks on these samples using MACS2.

I want to identify differential peaks across these three.

Questions:

a) How can I identify these peaks?

b) Is their any package available to do so, if yes, have you used it?

Kindly help.

Thank you

[Devon Ryan]

If you're hoping to get reliable p-values, you won't get any without replicates. You might just make a consensus peak set and then calculate pairwise fold-changes of normalized coverage between the conditions in them.

Answer 1

[Ming Tommy Tang]

I highly recommend you to try diffreps http://code.google.com/p/diffreps/

[ChIP]

Will it work without replicates?

[Ming Tommy Tang]

yes, it works without replicates, but you need to use G-test

Answer 2

[Ying W]

With no replicates, you might be best off just overlapping the peaks found and separating it into classes of overlaps (think venn diagram). You could also give DBChIP a try ( bioconductor link ) since it has an alternate way that tries to estimate variation w/out replicates.

Answer 3

[Song Qiang]

Please checkout the RSEG tools at http://smithlab.usc.edu/histone/rseg/ I developed this tool using Hidden Markov model. If you use the rseg-diff program with mode 3, it will identify regions with differential read counts.

Please let me know if you have any questions.

[ChIP]

Hi! Thanks for the link.

I installed every bit of thing it said and at the last step, the script fails to write down the output files.

I am using -o $PWD

Error says " cannot write it in the folder"

Could you kindly reply

Answer 4

[dnaseiseq]

These guidelines might be helpful:

Bailey T, Krajewski P, Ladunga I, Lefebvre C, Li Q, Liu T, Madrigal P, Taslim C, Zhang J (2013) Practical Guidelines for the Comprehensive Analysis of ChIP-seq Data. PLoS Comput Biol 9(11): e1003326. doi:10.1371/journal.pcbi.1003326

http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1003326