A new framework for computational protein design through cost function network optimization.

Area of Science:

• Computational Biology
• Biochemistry
• Bioinformatics

Background:

• Computational protein design (CPD) faces challenges with large, combinatorial search spaces, making exact exploration difficult.
• The NP-hard nature of CPD limits the analysis of complex sequence-conformation spaces.

Purpose of the Study:

• To introduce a novel CPD framework leveraging cost function network (CFN) solving for efficient combinatorial optimization.
• To demonstrate the framework's capability in handling complex protein design problems intractable for existing tools.

Main Methods:

• Utilized a cost function network (CFN) solving approach, an exact combinatorial optimization technique.
• Integrated energetic model generation (Osprey 2.0), CFN conversion (Perl scripts), and CFN solving (toulbar2).

Main Results:

• The CFN-based CPD framework successfully solved complex designs to optimality, outperforming standard CPD tools and operations research methods.
• The method efficiently identified global minimum-energy conformations and enumerated large ensembles of suboptimal solutions.

Conclusions:

• The CFN-based framework offers an effective solution for complex combinatorial optimization in protein design.
• This approach facilitates the rational design of enzyme mutant libraries by providing optimal and suboptimal conformational data.