coolpup.py

coolpup.py performs pile-up analysis of Hi-C data in the .cool format to average interaction matrices across genomic features and enhance detection of chromosomal loops and interactions.

Key Features:

• Input format: Operates on Hi-C contact matrices stored in the .cool file format.
• Genome-wide pile-up analysis: Performs genome-wide averaging (pile-ups) of peaks or other annotated features to aggregate interaction signals.
• Improved sensitivity for low-depth data: Enhances detection of genomic interactions in datasets with limited sequencing depth, including single-cell Hi-C scenarios.
• Performance optimization: Loads entire chromosomes into memory sequentially or in parallel and rapidly extracts small submatrices to enable large-scale analyses.
• Pile-up variation for statistics: Implements a variation of the pile-up approach to support statistical analysis of looping interactions.
• Biological validation: Has been used to reproduce published findings on cohesin and CTCF and to investigate Polycomb-driven interactions.

Scientific Applications:

• 3D genome organization: Analysis of chromosomal interactions and loop structures from Hi-C data.
• Low-depth and single-cell Hi-C studies: Aggregation-based analysis to enable interaction detection when sequencing depth is limited.
• Protein-mediated interaction studies: Investigation of roles of cohesin, CTCF, and Polycomb in genome architecture.
• Statistical evaluation of loops: Quantitative assessment of looping interactions using the tool's pile-up variation.

Methodology:

Performs genome-wide averaging (pile-ups) of annotated features on .cool Hi-C matrices by loading chromosomes into memory sequentially or in parallel, extracting small submatrices, and applying a variation of the pile-up approach for statistical analysis.