Integration of ATACseq data with ChIPseq data

[Anna Kuchmiy]

Hi all, I am experiencing some difficulties in the integration of ATACseq data with K27me3 ChIP data. Peaks were called with MACS2 --nomodel --extsize75 , broad peaks in case of ChIP data, for ATACseq there were narrow peaks. So after summarizeOverlaps counts and DeSEQ2 analysis for ATAC and K27me3 separately the significant regions of ATAC do not overlap with regions of K27me3. It might be a biological effect. However, a similar analysis of RNAseq and H3K4me3 overlap also didn't reveal overlap in significant regions. This is very weird because I would expect the regions to overlap like 90%. How to interpret such a result? Should I do summarizeOverlap counts with ATAC and ChipSeq altogether from the beginning?
Many thanks in advance,
Anna