X-ray vs. NMR structures as templates for computational protein design

Area of Science:

• Structural biology
• Computational chemistry
• Protein engineering

Background:

• Protein design calculations often utilize experimentally determined structures as templates to create novel protein sequences.
• High-resolution crystal structures have historically been favored for these template-based designs.
• The availability of nuclear magnetic resonance (NMR) structures prompts an investigation into their suitability as templates.

Purpose of the Study:

• To compare the efficacy of X-ray versus NMR structures as templates in protein design calculations.
• To assess differences in side-chain repacking and design outcomes using both template types.
• To determine if NMR templates can be a viable alternative when crystal structures are absent.

Main Methods:

• Assembled a database of 29 proteins with both high-resolution X-ray and NMR structures.
• Compared template quality using metrics like rotamericity and native-sequence recovery.
• Performed protein design calculations using RosettaDesign with both X-ray and NMR templates.
• Evaluated the energy and packing quality of the designed proteins.

Main Results:

• X-ray templates generally outperformed NMR templates in RosettaDesign calculations.
• For about 20% of proteins, NMR ensemble members yielded comparable design results.
• Re-evaluation with different energy functions revealed smaller differences between X-ray and NMR template performance.

Conclusions:

• While X-ray structures are often superior templates for protein design, NMR ensembles can be effective alternatives.
• The absence of a crystal structure should not deter computational design efforts if NMR data is available.
• Experimental validation remains crucial for confirming the utility of any chosen template structure.