HiCzin

HiCzin normalizes metagenomic Hi-C contact maps and detects spurious inter-species contacts to correct explicit and implicit biases and improve interpretation of microbial three-dimensional genome architecture.

Key Features:

• Normalization of Metagenomic Hi-C Data: Normalizes metagenomic Hi-C contact maps to enable more accurate downstream analyses of chromosomal interactions in microbial communities.
• Detection of Spurious Contacts: Identifies and removes spurious inter-species contacts using a zero-inflated negative binomial regression model.
• Parametric Modeling Approach: Employs a parametric modeling framework to model contact counts and correct systematic effects.
• Correction of Explicit and Implicit Biases: Addresses both explicit biases (systematic, directly identifiable) and implicit biases (subtle, requiring modeling) in metagenomic Hi-C data.
• Enhanced Metagenomic Contig Clustering: Improves metagenomic contig clustering performance through normalized and decontaminated Hi-C contact maps.

Scientific Applications:

• Spatial Genome Organization: Elucidates three-dimensional genomic architecture within microbial communities using normalized Hi-C contact maps.
• Genome Reconstruction and Contig Clustering: Supports metagenomic contig clustering and genome binning by providing bias-corrected contact information.
• Interaction Networks and Functional Relationships: Facilitates inference of interaction networks and functional relationships among species in complex microbial ecosystems.

Methodology:

HiCzin applies a parametric, zero-inflated negative binomial regression model to correct explicit and implicit biases in metagenomic Hi-C contact maps and to identify spurious inter-species contacts.