HiCHap 1.0

Introduction

HiCHap is a Python package designed to process and analyze Hi-C data, primarily for diploid Hi-C by using phased SNPs. First, the Hi-C reads are split in ligation junction sites, and then all split parts are used in mapping to maximumly utilize SNPs in allele assignment, thus improving the ratios of allele-assigned reads. The noisy reads are further eliminated. Second, except for traditional data bias caused by Hi-C experiments, the unevenly distributed genetic variants lead to additional bias in reconstructed Hi-C haplotype because it is potentially easier to assign allelic contacts in the chromatin regions with denser genetic variants. HiCHap utilizes a two-step strategy to reduce these two types of data biases by using the mapped and allele-assigned contacts only. Third, with the improved quality of reconstructed Hi-C haplotype, HiCHap can identify compartments, topological domains/boundaries and chromatin loops at haplotype level, and also provide testing on the allelic specificity for these structures. Finally, HiCHap supports data processing, bias correction and structural analysis for traditional Hi-C without separating homologous chromosomes.

Requirements

HiCHap is developed and tested on Unix systems. HiCHap utilizes HDF5 and cooler as default data format to keep consistent with 4DN standards. To summarize, the following packages are required in installation.

python packages:

1. Python 2.7+

2. Multiprocess

3. Numpy

4. Scipy

5. statsmodels

6. Scikit-Learn

7. xml

8. pysam

9. ghmm

10. Bio

11. cooler

others:

1. bowtie2 (version 2.2.9 is tested)

2. samtools (version 1.5 is tested)

Downloads

• Source code and manual here

• Code and manual Repository (At GitHub, Track the package issue)

Citation