search

XiaoTaoWang / HiCPeaks

A Python implementation for BH-FDR and HiCCUPS
GNU General Public License v3.0
41 stars 9 forks source link

Output coordinates question #2

Closed Phlya closed 5 years ago

Phlya commented 5 years ago

Hi, thank you very much for this tool, I have tried to use it and was pleasantly surprised - very easy to use and fast!

I however have a small question about the output coordinates of pyHICCUPS. What do they correspond to? What is the difference between loc_1 and centroid_x? Sometimes they are the same, and sometimes they are not... And how is radius determined?

Thank you, Ilya

Phlya commented 5 years ago

Basically, what is the correct way to convert the output to bedpe for visualization in e.g. HiGlass?

XiaoTaoWang commented 5 years ago

Hello, use loc_1 and loc_2 instead of centroid_x and centroid_y in post analysis and visualization, because (loc_1, loc_2) corresponds to the peak with the greatest interaction frequency in each cluster, while (centroid_x, centroid_y) just indicates the cluster centroid.

XiaoTaoWang commented 5 years ago

pyHICCUPS implements the greedy clustering algorithm proposed by the original paper (Rao et al. Cell. 2014), and the radius is defined as the Euclidean distance between the centroid of a cluster and the furthest peak pixel in the same cluster.

Phlya commented 5 years ago

Thank you very much for the explanation!