Positive multistate protein design.

Area of Science:

• Protein Engineering
• Computational Biology
• Biophysics

Background:

• Structure-based computational protein design (CPD) is crucial for protein engineering.
• Traditional CPD methods often overlook backbone flexibility by focusing on a single rigid structure.
• Multistate design (MSD) addresses this by considering multiple backbone conformations, posing significant computational challenges.

Purpose of the Study:

• To develop efficient computational methods for positive multistate protein design (MSD).
• To implement these methods in user-friendly software for broader accessibility.
• To evaluate the performance of different fitness definitions in sequence recovery for MSD.

Main Methods:

• Efficient reductions of positive MSD problems to Cost Function Networks were developed.
• Two distinct fitness definitions were employed within these networks.
• The developed algorithms were implemented in the open-source software Pompd (Positive Multistate Protein design).

Main Results:

• Pompd can identify guaranteed optimal sequences and enumerate suboptimal sequences for positive MSD problems.
• Application to NMR and X-ray structures showed that average energy fitness yields the best sequence recovery.
• The method significantly outperforms existing guaranteed computational design approaches, solving previously intractable problem sizes.

Conclusions:

• The developed method and Pompd software offer a significant advancement in computational protein design.
• Accounting for backbone flexibility via MSD is feasible and effective.
• Pompd enables solving larger and more complex multistate protein design problems than previously possible.