pyFOOMB

pyFOOMB implements object-oriented ordinary differential equation (ODE) models for quantitative characterization and parameter estimation of bioprocesses.

Key Features:

• Object-Oriented Modelling: Implements bioprocess models as systems of ODEs using an object-oriented design for modular model composition.
• Integration with Python Ecosystem: Integrates with existing Python packages to incorporate experimental data and support iterative workflows for parameter estimation.
• Replicate Model Instances: Employs replicate model instances linked by common parameters with global or local properties to represent multiple stages or varying conditions.
• Event Handling: Handles discontinuities in differential equations for multi-stage processes using the assimulo package.
• Optimization Capabilities: Supports optimization via a parallelized generalized island approach from the pygmo package to refine model parameters toward KPIs such as titer, rate, and yield.
• Parameter Estimation and Non-linear Regression: Enables parameter estimation and non-linear regression for quantitative determination of model parameters and performance indicators.

Scientific Applications:

• Quantitative characterization of bioprocesses: Provides quantitative model-based characterization of biotechnological production processes.
• Optimization of fermentation processes: Facilitates optimization of fermentation conditions and strategy through parameter refinement and objective-driven search.
• KPI determination: Supports determination of key performance indicators (titer, rate, yield) via parameter estimation and non-linear regression.
• Multi-stage process modelling: Enables modelling and analysis of multi-stage processes with stage-specific parameterization and event-driven transitions.

Methodology:

Constructs bioprocess models as ODE systems in an object-oriented manner, uses replicate instances with global/local parameters, handles events via assimulo, performs optimization with a parallelized generalized island approach from pygmo, and applies parameter estimation and non-linear regression while integrating experimental data through Python packages.