search

dozmorovlab / HiCcompare

Joint normalization of two Hi-C matrices, visualization and detection of differential chromatin interactions. See multiHiCcompare for the analysis of multiple Hi-C matrices
https://dozmorovlab.github.io/HiCcompare/
Other
18 stars 3 forks source link

Why are the results from method KRnorm different from results from Juicertools #23

Closed junmeiW closed 2 years ago

junmeiW commented 3 years ago
Hi, I've use the method KRnormand SCNto normalization. I found that results from these 2 methods are all less than 1, and the KRnorm results are totally different from that from Juicertools For example, results from KRnorm region1 region2 IF
3000000 3000000 0.315465
3000000 3100000 0.081327
3100000 3100000 0.317527
3000000 3200000 0.042387
and Juicertools region1 region2 IF
3000000 3000000 1816.6257
3000000 3100000 468.3274
3100000 3100000 1828.491
3000000 3200000 244.09053

Are these results reasonable?

Looking forward to your reply!

mdozmorov commented 3 years ago

It is possible there is a bug in our implementation of KRnormalization.R. The KR normalization supposed to make rows and columns sum to one. Try to find Juicer's code and compare it with our implementation, or, better yet, with the original KR paper.

junmeiW commented 3 years ago

Thanks to your suggestion, I'll check that.