Fig. 4

An example to define custom tracks that display Hi-C contact matrix along with peaks detected by Peakachu algorithm [17]. An example of peak prediction result is demonstrated in the right panel. The upper triangular matrix shows the peak p-value output by Peakachu algorithm. The predicted peaks drawn as blue squares upon the original matrix is shown in the lower triangular matrix. The Hi-C matrices and the peaks upon them will be automatically recomputed and updated after the change of genome region. The left panel is the full Runnable python codes used for generating the right panel. The custom track is implemented by following the same intuitive and clear design pattern as other built-in tracks: i.e., reusing the data fetching and plotting functionality as much as possible. For this figure, the functionality of fetching and plotting contact matrix with peaks are totally reused by inheriting Cool/ArcsBase track base class, and the rest of the codes merely calls the computing function of the peachachu package. After the track definition, we can see that the custom track is born to support being used in a ggplot2-like syntax with other tracks, and this capability is also valid in CLI and GUI mode