Introduction

3C-based techniques(5C, Hi-C) have revealed the existence of topologically associating domains(TADs), a pervasive sub-megabase scale structure of chromosome. TADs are contiguous regions in which loci interact much more frequently with each other than with loci out of the region. Visually, TADs appear as square blocks along the diagonal on a heatmap.

There are various methods for TAD identification [1], [2]. Most methods apply a two-step scheme: First, transform TAD or boundary signal into 1d profile using some statistic(e.g. Directionality Index, DI); Then, use the 1d profile to identify potential boundaries and produce a set of discrete non-overlapping TADs. However, the organization of chromosome structure is always intricate and hierarchical. Phillips-Cremins JE et al. [3] utilized a modified DI of multiple scales subdivided TADs into smaller subtopologies (sub-TADs) using 5C data. Here, I extend their algorithm to the whole genome and develop this software.

calTADs are tested on traditional [4] and in-situ [5] Hi-C data, both generating reasonable results.

Installation

Please check the file “INSTALL.rst” in the distribution.

Links

• Repository

• PyPI

Usage

Open a terminal, type calTADs -h for help information.

calTADs contains a process management system, so you can submit the same command repeatedly to utilize the parallel power as much as possible.

Reference