A generic framework for hierarchical de novo protein design

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• De novo protein design aims to create novel protein sequences and structures beyond those found in nature.
• Designing 'designable' structural templates is a key challenge for predicting protein folding.
• Understanding protein folding principles is crucial for developing proteins with new functions.

Purpose of the Study:

• To develop an automated framework for de novo protein design using structural templates.
• To eliminate the need for hand-crafted rules in protein design by using a data-driven approach.
• To generate artificial proteins with customized geometries for exploring the protein universe.

Main Methods:

• The TopoBuilder (TB) method was extended to automatically assemble structural templates from string descriptors of protein topology.
• A data-driven, iterative approach was employed, extracting geometrical parameters from tertiary motifs.
• The framework was evaluated by designing sequences for five distinct protein folds.

Main Results:

• The TopoBuilder framework successfully generated sequences for multiple protein folds.
• Experimental characterization confirmed that several designed sequences were folded and stable in solution.
• The method demonstrates an automated, template-based approach to de novo protein design.

Conclusions:

• The TopoBuilder framework provides a broadly useful tool for generating artificial proteins with specific geometries.
• This approach facilitates the exploration of novel protein structures and sequences.
• The study validates a data-driven method for de novo protein design, advancing the field of protein engineering.