How to locate the highly different compartment

[biozzq]

Dear all,

How can we locate the highly different compartment quickly, just by the lowest padj? Maybe you can explain the parameters for us, such as mdist and dZsc, which can help us to understand the results. Thank you.

Best wishes, Zheng Zhuqing

[ay-lab]

Hello,

Apologies for the lack of documentation on this subject. A pre-print is forthcoming with full descriptions of m-distance and dZsc; they are other multivariate statistical measures used for the final log10 PAdj value. For identifying differential compartments, however, use the Log10 PAdj values (dark green).

[biozzq]

Dear @ay-lab

I can not detect any A/B compartment transitions, most signals show variability in A/B compartment scores, like following screenshot. Maybe these results indicate the global folding patterns of chromosomes are highly similar between case and control?

Best wishes, Zheng zhuqing

[ay-lab]

Looking at this screenshot, it does appear that the compartments for your two runs are very similar (it almost seems like they are replicates); checking other chromosomes may confirm that.

[biozzq]

Dear @ay-lab

Thank you. The other chromosomes are same with above screenshot. Thus, I need treat this as the variability in A/B compartment scores rather than A/B compartment transitions. Is this OK?

Best wishes, zheng zhuqing

[ay-lab]

Yes—MFA normalization ensures that the scaling etc is normalized and the values are comparable

[biozzq]

Dear @ay-lab

Thank you for your help.

Best wishes, Zheng zhuqing