Scripts for prediction of long-range interactions between regions/domains based on Hi-C maps. Establish contact!
HiCEnterprise takes a list of regions/domains and searches for their significant interactions on given Hi-C maps. Package consists of two types of analyses: regions and domains.
This part of the package is an implementation of the method for identification of long-range interacting regions and creating interaction profiles based on Hi-C data as introduced in (Won, 2016).
The significance of the individual pairwise interaction between bins is calculated as the probability of observing a stronger contact under the fitted Weibull distribution matched by chromosome and distance. FDR (False Discovery Rate) is used for correcting for multiple testing.
In the domain part of the package the significance of interaction between domains A and B is calculated as the probability of observing a stronger contact under the fitted Hypergeometric distribution, where N (population size) is the sum of all frequencies in the Hi-C map, a is the sum from the column of domain A and b is the sum from the row of domain B (number of success states in the population and number of draws).
HiCEnterprise type options
type could be either regions or domains, options are type dependent.
If you want to see the list of available arguments with short explanations type either HiCEnterprise regions --help
or HiCEnterprise domains --help
. Full specifications for each argument can be found in the documentation (docs folder or online).
python setup.py install
You may need administrative privileges to do this.
To test if the scripts behave in a way they should be find the main HiCEnterprise directory and run:
python setup.py test
pytest package is needed for running the test framework.
We provided some example files (used for testing too) with which you can learn to use the program. Here you can see two example runs:
HiCEnterprise regions
HiCEnterprise domains
Additional examples and specifications for each argument can be found in the documentation (docs folder or online).
Hania Kranas, 2017