ModelCIF Update: Supporting Emerging Classes of Computational Macromolecular Models

Area of Science:

• Computational structural biology
• Bioinformatics
• Data standards

Background:

• Highly accurate protein structure prediction tools have expanded computational structural biology.
• Existing ModelCIF (1.4.4) primarily supported homology and ab initio modeling.
• Need for standardized data representation for new modeling use cases.

Purpose of the Study:

• Extend the ModelCIF data standard to support new computational modeling use cases.
• Define guidelines for storing and validating modeling results.
• Update associated tools for model generation, validation, archiving, and visualization.

Main Methods:

• Extended ModelCIF data standard with new categories.
• Defined guidelines for capturing protocols, inputs, and outputs for new use cases.
• Updated software tools to implement the extended standard.

Main Results:

• ModelCIF extension now supports protein-ligand and protein-protein interactions, conformational sampling, and de novo protein design.
• Established guidelines for storage and validation of these new modeling results.
• Enhanced software tools facilitate model generation, validation, archiving, and visualization.

Conclusions:

• The extended ModelCIF framework promotes FAIR data principles for computational models.
• Improved reproducibility and reusability of modeling results in life sciences.
• Facilitates broader applications in medicine, drug discovery, and engineering.