pyPESTO

pyPESTO performs parameter estimation for quantitative dynamical models by inferring unknown model parameters from qualitative (categorical) and quantitative experimental data.

Key Features:

• Optimal Scaling Method Integration: Incorporates an optimal scaling method from statistics to transform categorical (qualitative) data into quantitative representations while respecting relational constraints.
• Reduced Formulation for Optimization: Implements a reduced optimization formulation that requires fewer degrees of freedom while preserving the same optimal points as the full formulation.
• Improved Robustness and Convergence: Uses the reduced formulation to improve robustness and convergence properties of optimizers in parameter estimation tasks.
• Reduced Computation Times: Enhancements in optimizer performance lead to substantially reduced computation times for parameter estimation on complex models.

Scientific Applications:

• Systems Biology Model Parameterization: Enables parameter estimation for quantitative dynamical models commonly used in systems biology and related disciplines.
• Parameter Estimation from Qualitative Data: Facilitates incorporation of qualitative (categorical) experimental data into parameter inference workflows.
• Model Refinement and Prediction: Supports refinement of model parameters to improve model accuracy and predictive power for experimental design and intervention analysis.

Methodology:

Transforms qualitative variables into quantitative values via optimal scaling and solves a reduced optimization problem that preserves optimal points to estimate model parameters.