How to filter the results from `hicDifferentialTAD` based in p-value and W values?

[Linhua-Sun]

Hi: I found hicDifferentialTAD is quite useful to get differential interactions domains from two matrix. However, I am wondering how to filter the results. The default outputted table is quite confuse. Since there are 5493 rejected TADs and 292 accepted TADs. Such a result is obviously unrealistic based on visual inspection. I have downsampled two cooler into similar valid pairs and further KR normalized. My command line is hicDifferentialTAD --targetMatrix Mutants_Sub_1k_ToSmallKR.h5 --controlMatrix Control_Sub_1k_ToSmallKR.h5 --tadDomains MM_WT_Sub_1k_ToSmallKR_domains.bed --outFileNamePrefix KR_Small_Test. For example, if I focused on the type of intra-TAD. I ploted the distribution of p-value and W values. I'm not sure how to filter the table based on p-value and W. Or is there a problem in the process?

I used HiCExplorer version 3.7.2 with python 3.7.12.

Thanks for your time!

[joachimwolff]

Hi,

• have you normalised the data of the two matrices and if, how?
• can you please provided a plot of a specific region with the TADs as an example?
• you can also change the p-value that is used to decide acceptance / rejection
• Have in mind H0 is 'regions are equal' and 'acceptance' means this condition is fulfilled.
• What is your expectation? Shall there be major differences in the TADs or are the samples more or less equal?

Best,

Joachim