LikelihoodProfiler

LikelihoodProfiler performs identifiability analysis and estimates parameter confidence intervals for systems biology and Quantitative Systems Pharmacology (QSP) models.

Key Features:

• Identifiability Analysis: Determines how accurately model parameters can be recovered from experimental data to assess practical identifiability.
• Confidence Intervals Evaluation: Computes parameter confidence intervals using profile likelihood methods.
• CICO Algorithm: Implements the Confidence Intervals by Constraint Optimization (CICO) algorithm to accelerate and improve confidence interval estimation.
• Customizable Accuracy Settings: Exposes adjustable numerical settings in the CICO implementation to control the precision of confidence interval estimates.

Scientific Applications:

• Systems biology modeling: Applied to dynamical systems biology models to quantify parameter identifiability and uncertainty.
• Quantitative Systems Pharmacology (QSP): Used to evaluate parameter identifiability and confidence intervals in QSP models.
• Taxol treatment models: Applied to models related to Taxol treatment to assess parameter confidence and identifiability.
• STAT5 Dimerization models: Applied to STAT5 Dimerization models to evaluate parameter identifiability and confidence intervals.

Methodology:

Uses the profile likelihood approach and integrates the Confidence Intervals by Constraint Optimization (CICO) algorithm, with a numerical implementation that includes adjustable precision settings.