how to make HPeak work for non-human ChIPSeq and split out p-values ??

[yuelics]

HPeak is a very nice HMM-model-based peak calling program. However, the current version HPeak 2.1 has two big problems: 1. only works for human genome; 2. output containing no p-value.

I wonder if anyone came across these two problems before and found a way around them.

Thanks in advance.

[peterbengkui]

Hey there,

HPeak now has been updated and you can find the newest version at:
https://sourceforge.net/projects/hpe...atest/download.

Sorry I just the post has been over a year old....

[yuelics]

Thanks Peter! I downloaded the package and see 3 new scripts: chiphmm_mm9 hmmminus_mm9 peakann_mm9.pl

did you how to make HPeak to run on the "mm9" but not the "hg19" scripts? From the manual, there does not seem to have an option to specify that:

perl HPeak.pl -format ELAND -t example.txt -n output –fmin 100 - fmax 300 -w 25 -s 1e-3 -wig -seq -ann

[yuelics]

Sorry I found it ... it is the "-sp" option in the HPeak.pl script

[peterbengkui]

Yue,

can you give me some feedback for your mouse data analysis? The HMM algorithm used might not be the best one.

You can contact me anytime at: [email protected]

Best,
Peter