HRM

HRM models stochastic transcription–translation gene regulatory networks using a hierarchical statistical inference framework for parameter estimation and mechanistic analysis in systems biology.

Key Features:

• Stochastic Modeling: Represents gene regulatory networks via a coarse-grained model with stochastic (binary) promoter states coupled to continuous protein variables.
• Statistical Inference Framework: Implements an exact inference algorithm and a variational approximation to enable scalable parameter inference and learning.
• Scalability and Flexibility: Combines exact and approximate inference techniques to scale calibration to experimental data and to analyze networks of small to medium scale.
• Mechanistic Detail: Models transcription–translation interactions through coupled non-linear ordinary differential equations (ODEs) while retaining a coarse-grained representation.

Scientific Applications:

• Gene Regulatory Network Analysis: Study dynamic behavior of gene regulatory networks, capturing stochastic promoter switching and protein dynamics.
• Parameter Estimation and Model Calibration: Infer model parameters from experimental data using exact and variational inference for accurate calibration.
• Novel Biological Predictions: Generate testable predictions about gene regulatory mechanisms from inferred models and case studies.

Methodology:

Coarse-grained modeling with binary promoter states and continuous protein variables; representation of interactions by coupled non-linear ordinary differential equations (ODEs); exact inference algorithm; variational approximation for scalable inference.