Skip to content

Latest commit

 

History

History
56 lines (37 loc) · 2.58 KB

README.md

File metadata and controls

56 lines (37 loc) · 2.58 KB
Raw

Calculate and visualize the contact map

Calculate contact maps

We provide a python script calContactMap.py to calculate the contact map from simulated chromatin structures. It produces a txt file in the ./cmap/ folder to store the contact matrix. When multiple simulations are performed, the contact map for each individual simulation will be stored in the folder ./[Celltype]/[chrId]/[runId]/.

Usage:

python calContactMap.py --Cell <Celltype> --njob <number_of_jobs> --job <jobname> --user <username> --ptn <partition> --chrom <chromosome_id_array>

or default:

python calContactMap.py

Note items in [] are optional. By default is calculating: Gm12878, chromosome 1, 8 parallel running.

[--Cell] can be selected from the following list (case sensitive):

Gm12878
H1hesc
Hela
Hepg2
Huvec
K562

[--njob] specifies the number of parallel running. By default, the value is 8.

[--job] specifies the name of the job on the cluster.

[--user] [--ptn] should be specified based on the cluster account and available cluster partition.

[--chrom] can be any non-repeated subset selected from 1 to 22. The accepted format is:

1 10 19 21

The manual would be available by executing:

python calContactMap.py -h

Note that the python script calls a FORTRAN code to perform the actual calculation of contact maps. The code has been precompiled with ifort compiler. One can also modify the script to compile with gfortan if ifort is not available.

Visualize contact maps with MATLAB

To visualize the contact maps, a MATLAB script visContactMap.m is provided. Note that the corresponding raw Hi-C maps with consistent resolution need to be downloaded beforehand by executing the following command:

cd ./hic/rawMap/; ./download.sh

A pack file for the contact map of GM12878 would be downloaded and then unpacked, which may take a little while for the whole process. The raw Hi-C maps are downloaded from:

Rao, Suhas S.P. et al. Cell 159, 1665-1680 (2014).

Then open MATLAB and execute the script visContactMap.m to plot the contact map. Type in the parameters such as the cell type, chromosome id and genomic positions accordingly. Leave the option empty if the default is satisfied. Note that the path for the Hi-C maps will be directed to the location where the Hi-C maps are downloaded by default when executing the MATLAB script.