InfMod3DGen

InfMod3DGen infers ensembles of three-dimensional chromatin conformations from chromosome conformation capture (3C)-based interaction data using a Bayesian framework that integrates polymer physics priors.

Key Features:

• Bayesian Framework: Employs a Bayesian framework that incorporates prior knowledge of polymer physics and conformational energy into chromatin structure inference.
• Polymer Chain Modeling: Models each chromosome as a polymer chain to capture spatial folding and long-range chromatin interactions.
• Expectation-Maximization Algorithm: Uses an expectation-maximization (EM) algorithm to estimate unknown model parameters and infer an ensemble of structures from 3C-based interaction frequency data.
• Validation and Verification: Applies cross-validation and verification against geometric constraints from fluorescence in situ hybridization (FISH) and known genetic interactions reported in the literature.
• Comprehensive Analysis: Computes ensembles of 3D chromatin conformations that reconcile distance constraints from 3C-based data with other experimental evidence.

Scientific Applications:

• 3D Genome Organization: Provides inferred chromatin ensembles for studying spatial arrangements of chromosomes within eukaryotic nuclei.
• Gene Regulation and Differentiation: Enables analysis of long-range chromatin interactions relevant to gene regulation and cellular differentiation.
• Disease Genomics: Facilitates investigation of disease pathogenesis related to genomic architecture by linking 3D structure to genetic interactions.

Methodology:

Implements a Bayesian statistical model integrating polymer physics priors, represents chromosomes as polymer chains, estimates parameters via an expectation-maximization (EM) algorithm, and evaluates results with cross-validation.