Towards optimal virtual patients: an online adaptive control approach

Area of Science:

• Computational biology
• Pharmacology
• Biomedical engineering

Background:

• In silico disease models enhance drug discovery by simulating patient heterogeneity.
• Diverse virtual patient cohorts improve the robustness of computational analyses and hypothesis testing.
• Manually tuning parameters for complex biological systems is infeasible for creating virtual patient populations.

Purpose of the Study:

• To present an online adaptive control technique for optimizing virtual patient model parameters.
• To automatically generate a diverse virtual patient cohort that meets specific stimulus-response constraints.
• To enhance biosimulation applications in drug discovery and optimization.

Main Methods:

• Model Reference Adaptive Control (MRAC) technique for parameter auto-tuning.
• Application of the control scheme on the Entelos Metabolism PhysioLab platform.
• Validation of the automated generation of virtual patient cohorts.

Main Results:

• Successfully implemented an online adaptive control technique for parameter optimization.
• Automatically generated a validated cohort of virtual patients with desired behaviors.
• Demonstrated the efficacy of the MRAC-based approach for creating diverse in silico populations.

Conclusions:

• The proposed adaptive control technique effectively auto-tunes model parameters for virtual patient populations.
• This method facilitates the rapid development of validated virtual patients for in silico research.
• The approach enhances the application of biosimulation in drug discovery and optimization by addressing model complexity.