Related Experiment Video
Updated: Jun 19, 2026
10:24
Generation of Heterogeneous Drug Gradients Across Cancer Populations on a Microfluidic Evolution Accelerator for Real-Time Observation
Published on: September 19, 2019
Developing optimal input design strategies in cancer systems biology with applications to microfluidic device
Filippo Menolascina1, Domenico Bellomo, Thomas Maiwald
1Department of Electrical Engineering and Electronics, Technical University of Bari, Via E. Orabona 4, Bari, Italy. f.menolascina@ieee.org
BMC Bioinformatics
|October 16, 2009
Summary
Optimizing stimulus response experiments using evolutionary algorithms and Fisher Information Matrix enhances data acquisition for cancer systems biology. This approach minimizes model uncertainty and aids in understanding complex diseases.
Area of Science:
- Systems Biology
- Oncology
- Biochemical Engineering
Background:
- Mechanistic models are crucial in Systems Biology for understanding cancer pathways, but data scarcity hinders progress.
- Stimulus Response Experiments (SRE) provide insights into cellular mechanisms, yet optimizing input profiles for information retrieval remains challenging.
- Acquiring data for complex biological systems is expensive and time-consuming, exacerbating the 'data rich-data poor' paradox.
Purpose of the Study:
- To develop an optimal experimental design strategy for enhanced information gathering in Systems Biology.
- To quantify experimental information using the Fisher Information Matrix.
- To address the challenge of optimizing input time-profiles for stimulus response experiments.
Main Methods:
- Utilized the Fisher Information Matrix to quantify experimental information.
- Developed an optimal experimental design strategy employing an evolutionary algorithm.
- Investigated dynamical properties of cell stimulation signals based on control systems theory.
- Developed a microfluidic device for automated cell stimulation and system identification.
Main Results:
- The proposed approach was applied to the Epidermal Growth Factor Receptor pathway.
- Parametric uncertainty in the identified model was minimized.
- Optimally designed experiments demonstrated superiority over canonical inputs, confirmed by Monte-Carlo simulations.
- A microfluidic platform was developed for automated biochemical model identification.
Conclusions:
- The developed method effectively minimizes parametric uncertainty in biological models.
- Optimized experimental designs significantly improve information retrieval compared to standard methods.
- The approach is extensible to multiobjective formulations and identifiability analysis.
- Automated microfluidic platforms can alleviate data limitations in Systems Biology.
