NLoed: A Python Package for Nonlinear Optimal Experimental Design in Systems Biology

Area of Science:

• Systems and synthetic biology
• Computational biology
• Biotechnology

Background:

• Accurate modeling in systems biology requires precise parameter estimates, which depend on high-quality, relevant experimental data.
• Current optimal experimental design (OED) software often struggles with the complex, nonlinear models and diverse data types common in biological research.
• Existing OED tools lack integration with modern numerical methods, leading to computational inefficiencies.

Purpose of the Study:

• To introduce NLoed, an open-source Python library designed to enhance optimal experimental design (OED) for systems biology research.
• To provide a flexible and accessible tool for designing experiments that yield the most informative data for biological models.
• To address the limitations of current OED software in handling nonlinear models and diverse data types.

Main Methods:

• Development of NLoed, a modular Python library offering convenient access to OED methods.
• Implementation of support for a wide range of nonlinear, dynamic, multi-input/output models and various data types.
• Integration of maximum likelihood fitting and diagnostic tools for a comprehensive modeling workflow.

Main Results:

• NLoed provides a user-friendly interface for applying state-of-the-art OED techniques to biological modeling tasks.
• The software efficiently handles complex models and diverse data, improving the process of experimental design.
• Demonstrated successful application of NLoed for characterizing an optogenetic system in *Escherichia coli*.

Conclusions:

• NLoed offers a powerful, flexible, and accessible solution for optimal experimental design in systems biology.
• The software facilitates more efficient data collection and model calibration, advancing biological research.
• NLoed's capabilities are well-suited for the diverse experimental scenarios encountered in modern biological research.