SpaceScanner: COPASI wrapper for automated management of global stochastic optimization experiments

Area of Science:

• Biochemical Engineering
• Computational Biology
• Systems Biology

Background:

• Global stochastic optimization methods are widely used for biochemical network design.
• These methods have limitations including no guaranteed global optima, undefined termination criteria, and inability to detect stagnation.
• Manual intervention is often required to overcome these drawbacks, especially in large solution spaces.

Purpose of the Study:

• To develop an automated tool, SpaceScanner, to address the limitations of global stochastic optimization methods.
• To enhance the efficiency and reliability of biochemical network optimization.
• To facilitate the exploration of large parameter spaces in biochemical network design.

Main Methods:

• SpaceScanner employs parallel optimization runs to achieve automatic termination based on consensus.
• It implements a strategy to consecutively apply a set of global stochastic optimization methods when stagnation occurs.
• The tool supports automatic scanning of parameter subsets for optimal objective function values.

Main Results:

• SpaceScanner provides automatic termination criteria for optimization tasks.
• It mitigates stagnation issues by switching between optimization methods.
• A summary file of ranked solutions enables efficient analysis of parameter combinations.

Conclusions:

• SpaceScanner offers an automated and efficient solution for biochemical network optimization.
• The tool overcomes key limitations of traditional global stochastic optimization methods.
• It facilitates the design and improvement of complex biochemical networks by managing large parameter spaces effectively.