ASHIC

ASHIC models allele-specific chromatin contacts and infers allele-specific 3D chromatin structures from diploid Hi-C data using a hierarchical Bayesian framework.


Key Features:

  • Hierarchical Bayesian Framework: Uses a hierarchical Bayesian statistical model to infer allele-specific contact maps and allelic 3D structures from diploid Hi-C data.
  • ASHIC-PM (Poisson-Multinomial Model): Implements a Poisson-multinomial likelihood to model variability in Hi-C contact counts for imputing allele-specific contacts.
  • ASHIC-ZIPM (Zero-Inflated Poisson-Multinomial Model): Extends ASHIC-PM with zero-inflation to account for excess zeros in Hi-C data and improve fine-resolution contact imputation.
  • Allele-specific Contact Map Imputation: Imputes allele-specific contact matrices from diploid Hi-C data to separate homologous chromosome signals.
  • Allelic 3D Structure Inference: Simultaneously infers allele-specific 3D chromatin structures from the imputed allele-specific contact matrices.
  • Robustness to Low Coverage and Low SNP Density: Demonstrates improved performance in simulation studies under conditions of low sequencing coverage and low SNP density.
  • Fine-Resolution Chromatin Mapping: Produces fine-resolution chromatin maps, particularly when using the ZIPM model to handle sparse contacts.

Scientific Applications:

  • Fine-Resolution Chromatin Mapping: Generation of detailed allele-specific chromatin contact maps at high resolution.
  • 3D Structural Inference: Reconstruction of allelic 3D chromatin structures to study spatial genome organization.
  • Allelic Chromatin Organization Analysis: Analysis of allele-specific chromatin conformation in diploid Hi-C datasets from mouse and human samples.
  • Genetic and Functional Studies: Investigation of genetic diversity, disease mechanisms, and evolutionary biology through allele-specific chromatin architecture.

Methodology:

Imputes allele-specific contact maps from diploid Hi-C data using hierarchical Bayesian models (Poisson-multinomial and zero-inflated Poisson-multinomial) to distinguish homologous chromosomes and uses the resulting allele-specific contact matrices to infer 3D chromatin structures.

Topics

Details

Tool Type:
command-line tool
Programming Languages:
Python
Added:
1/18/2021
Last Updated:
11/24/2024

Operations

Publications

Ye T, Ma W. ASHIC: Hierarchical Bayesian modeling of diploid chromatin contacts and structures. Unknown Journal. 2020. doi:10.1101/2020.08.29.273722.

Ye T, Ma W. ASHIC: hierarchical Bayesian modeling of diploid chromatin contacts and structures. Nucleic Acids Research. 2020;48(21):e123-e123. doi:10.1093/nar/gkaa872. PMID:33074315. PMCID:PMC7708071.

PMID: 33074315
PMCID: PMC7708071
Funding: - National Science Foundation: DBI-1751317