Energy strain in three-dimensional protein structures

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Steric strain in protein 3D structures arises from unfavorable inter-atomic interactions.
• This strain can stem from packing, functional needs, or structural coordinate errors.
• Traditional energy functions are often too noisy for reliable error detection.

Purpose of the Study:

• To develop a sensitive method for detecting errors in protein structures using energy functions.
• To analyze the energy distribution of amino acid residues in high-resolution protein models.
• To validate the approach across various protein structure origins.

Main Methods:

• Applied a full-atom, detailed energy function after short energy refinement.
• Calculated statistics of amino acid residue energy distributions in crystal structures.
• Normalized residue energy deviations to identify strained, potentially incorrect regions.
• Compared structures from X-ray crystallography, NMR, theoretical models, and decoys.

Main Results:

• A refined detailed energy function effectively detects errors in protein structures.
• Normalized energy deviations pinpoint strained polypeptide chain fragments.
• The method demonstrated sensitivity across diverse protein structure types.
• Applications shown in loop and homology modeling.

Conclusions:

• Elevated energy strain indicates problematic protein structure fragments, whether experimental or theoretical.
• This approach aids in model building, refinement, homology modeling, protein design, folding calculations, and structure analysis.