A call for virtual experiments: accelerating the scientific process.

Area of Science:

• Computational Biology
• Systems Biology
• Scientific Methodology

Background:

• Experimentation is crucial for the scientific method, driving exploration, description, and explanation.
• Computational models are increasingly used but can lack usefulness and relevance.
• Virtual experiments (in silico analogues of wet-lab/field experiments) offer a path to improve models.

Purpose of the Study:

• To advocate for the reuse of virtual experiments to enhance computational models.
• To highlight the benefits of reusable virtual experiments for scientific research and systems biology.
• To propose a framework for integrating models and experimental data.

Main Methods:

• Reviewing the benefits of reusable virtual experiments.
• Proposing separate representation and annotation of models and experimental protocols.
• Outlining a method for confronting experimental datasets with candidate models.

Main Results:

• Reusable virtual experiments improve model specification, assaying, and comparison.
• They are essential for reproducible research, model reuse, and quality assurance.
• Separate, annotated representation of models and data facilitates linking.
• Rigorous model-data confrontation enables continuous integration of biological knowledge.

Conclusions:

• Promoting reusable virtual experiments significantly enhances the utility and relevance of computational models.
• This approach fosters critical scrutiny, serves as a common language, and is key to reproducible research.
• Integrating models with experimental data via a "continuous integration" process will transform systems biology.