Skip to content

Latest commit

 

History

History
52 lines (25 loc) · 3.66 KB

README.md

File metadata and controls

52 lines (25 loc) · 3.66 KB
Raw

trans-C

Identification of trans interacting DNA domains

Trans-C is a tool that finds in Hi-C data biologically meaningful sets of loci that interact strongly in trans.

It initiates probabilistic random walks with restarts from a set of seed loci to traverse an input Hi-C contact map and returns dense subnetwork containing the seed loci.

You can find more about the theory behind trans-C by reading our paper. If you use Trans-C in an academic setting please cite us.

Alt text

Dependencies

Trans-C is implemented in Python and uses the common numpy, sklearn, and scipy packages. Additionally, it relies on the iced and cooler packages, so make sure to install these dependencies (pip or conda install work effortlessly). We recommend using the conda tool for package and environment management.

Inputs

Trans-C has five required inputs:

  1. Hi-C matrix. This is the matrix that captures the genomic loci interactions and the main input to the pipeline. Accepted are the most commonly used formats: cooler (.cool), hic file (.hic), or numpy object (.npy). You can find many Hi-C datasets in the 4DN Data Portal: 4DN. In the rare chance you have your data in a different format (i.e .pairs or something else) you will need to convert it manually to one the three formats above.

  2. chrom_sizes. This is a simple tab delimited file that lists the chromosome names and their sizes. For example, you can take a look at this file listing the human chromosomes. These standard files are easily available for many organisms from the UCSC repository.

  3. bin_size. Integer specifying the resolution of the Hi-C matrix, that is the number of base pairs each bin has, for example 100,000.

  4. seed_file.bed. The files contains the location of the seed loci of interest to the researcher around which *trans-*clique will be built. trans-C will initiate random walks from these seed nodes and find other loci that interact with them in-trans. In our study we used five seed nodes that are related to a common biological function we examined, i.e the RBM20 splicing factory, or the five loci with most ChIP-seq peaks for a given transcription factor. Each loci should be listed in single line following the standard .bed file format.

  5. out_dir. Name of the directory the results should be written to.

Optionally, the user can specify:

  1. alpha. (default=0.5, must be between 0 and 1) This is the restart parameter that controls how far the random walk goes before it is forced to restart from a seed loci. Larger alpha forces frequent restarts and keeps the exploration of the interaction graph short, thus highlighting loci that are already strongly related to the seed loci. Smaller alpha allows for longer walks that explore further into the graph, hence increasing the importance of the global topology of the graph in relation to the seed loci.

How to run

The user needs to provide five required inputs as described above. These inputs need to be specified in a config.yml or passed as a command line arguments.

Then, simply run:

%> python bin/trans-C.py config.yml

or

%> python bin/trans-C.py hic_file(.cool/.hic/.npy) chrom_sizes.txt 100000 seed_file.bed out_dir (alpha 0.5)

Questions

Feel free to email Borislav Hristov: borislav at uw.edu