Strange results with peakcall

[moshe-olshansky]

Hello,
I have two samples, MT_029 and MT_030, in (MACS) BEDPE format.
I first called peaks for each sample with
callpeak -f BEDPE -g hs --keep-dup auto -t MT_029_sorted.bed -n 29
and similarly for sample MT_030.
Then I called peaks in MT_029 vs MT_030 and MT_030 vs MT_029 by:
callpeak -f BEDPE -g hs --keep-dup auto -t MT_030_sorted.bed -c MT_029_sorted.bed -n 30_vs_29
and similarly the other way.

What stroke me was that for individual peaks files about 10% of the peaks are in chr1, while ~46% of peaks in MT_029 vs MT_030 and ~60%(!) in MT_030 vs MT_029 are in chr1.

Could it be true?

Thank you,
Moshe Olshansky
Baker Heart & Diabetes Institute
Melbourne, Australia.