deDoc2

Introduction

deDoc2 is a TAD-like domain(TLD) prediction tool using structural information theory, it treats the Hi-C contact map as a weighted graph, and applys dynamic programming algorith to globally optimize the two-dimensional structural entropy of the graph partiton. The deDoc2 package consists of two predictors, deDoc2.w and deDoc2.s, to predict higher level, larger and lower level, smaller TLDs, respectively. The deDoc2.w minimizes the structural entropy in the whole Hi-C contact map, while the deDoc2.s minimizes the structural entropy in the matrices of sliding windows along the genome. We previously developed deDoc for bulk Hi-C TAD predicting at https://github.com/yinxc/structural-information-minimisation.

Usage

Download the jar package deDoc2.jar, or you can also use the source code.

Basic usage

deDoc2 relys on java environment.

java -jar deDoc2.jar [-inputfile file] [-sparseformat] [-binsize size] [-outputfile file] [-windowsize size] [-maxTLDsize size] [-rwr] [-rp rp] [-NDI] [-help]

>> java -jar deDoc2.jar -inputfile ./example/14.clean.con.chr18.matrix -binsize 40

Input Hi-C format

deDoc2 supports both contact matrix format and sparse format as in deDoc. deDoc2 needs no normalization process, raw Hi-C data is suitable.

• contact matrix format: n × n intra-chromosome contact matrix, tab split.

• sparse format: the first line is the number of bins, following lines are edges formed as from_bin to_bin edge_weight, space split.

1952
1 1 2.0
2 2 2.0
2 3 3.0
...

Parameters

Parameters	Required/Optional	Description	Default
-inputfile	Required	input Hi-C data file path	N/A
-sparseformat	Optional	whether the input data is stored as sparse format	N/A
-binsize	Required	binsize of the input data (kb)	40kb
-outputfile	Optional	output file path and prefix	the same as inputfile
-windowsize	Optional	size of the sliding window in deDoc2.s (Mb)	10Mb
-maxTLDsize	Optional	max size of the predicted TLD (Mb), used to speed up deDoc2	10Mb
-rwr	Optional	whether to perform RWR imputation before TLD prediction	N/A
-rp	Optional	the restart probability of RWR imputation	0.5
-NDI	Optional	whether to calculate normalized decoding information (NDI), which can be used to determine binsize	N/A
-help	Optional	print help information	N/A

Output files

• file suffixed .TAD is the result of deDoc2.w, the higher level of TLD.
• file suffixed .window.TAD is the result of deDoc2.s, the lower level of TLD.

The output files contains partitions as TLD results. Each line represents a TLD, containg the node id of this TLD, space split, numbering from 1. A single bin in one line should be removed as non-TAD regions.