From FAIR to CURE: Guidelines for Computational Models of Biological Systems

Area of Science:

• Computational biology
• Biomedical modeling
• Systems biology

Background:

• Scientific data management follows FAIR principles (Findable, Accessible, Interoperable, Reusable).
• Computational models are crucial for scientific progress, particularly in biology and medicine.
• Existing data guidelines do not fully address the unique needs of scientific models.

Purpose of the Study:

• To propose the CURE principles (Credible, Understandable, Reproducible, Extensible) as complementary guidelines for scientific models.
• To enhance the trustworthiness and impact of computational models, especially in biomedical applications.
• To advocate for a more disciplined approach to scientific modeling.

Main Methods:

• Defined the four CURE principles: Credible, Understandable, Reproducible, and Extensible.
• Detailed requirements for each CURE principle, including verification, validation, uncertainty quantification, clarity of descriptions, adherence to standards, and use of open standards.
• Outlined recommended and baseline requirements for implementing the CURE principles.

Main Results:

• The CURE principles provide a framework for managing scientific models.
• Specific recommendations are provided for ensuring model credibility, understandability, reproducibility, and extensibility.
• The proposed guidelines aim to improve the impact and trustworthiness of computational models.

Conclusions:

• Computational models require dedicated management guidelines, such as the CURE principles.
• Adoption of CURE principles will enhance the reliability and reusability of scientific models.
• Automation of CURE guidelines implementation is recommended to facilitate community adoption.