why there is significant difference in chicago output , when valid pairs are used from HiC Pro and HiCUP and which one is reliable ? and why chicago output also gives trans interaction as significant ?

[Former user]

Hi,

I am a novice to capture hic analysis. We found that when we use hicpro or hicup for calling valid pairs from capture data, we found significant difference in output we get for chicago. can you explain why this should happen and what is more reliable method to move ahead. ?

[Mikhail Spivakov]

Unfortunately, I am not familiar with HiC Pro - presumably, the count matrices are different between the two HiC alignment packages. We have always used HiCUP. Trans-interactions can be called as significant by design under very stringent criteria. Whether or not you wish to use them in downstream analyses depends on your specific question.