Determinants of side chain conformational preferences in protein structures

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Accurate prediction of protein side chain conformations is crucial for understanding protein structure-function relationships and for protein design.
• Existing methods for side chain rotamer prediction face challenges in accuracy and computational efficiency.

Purpose of the Study:

• To develop and evaluate a novel discriminatory function for selecting protein side chain rotamers based on statistical analysis of atomic contacts.
• To compare the performance of this method using local main chain, entire main chain, and simultaneous side chain pair construction.

Main Methods:

• Utilized a discriminatory function derived from statistical analysis of atomic contacts within protein structures.
• Evaluated side chain conformation prediction using local main chain, global main chain, and simultaneous pairwise side chain construction.
• Compared results against established side chain building methodologies.

Main Results:

• The method achieved approximately 75% accuracy for chi1 angles within 30 degrees of experimental values when using only the local main chain.
• An average side chain atom root-mean-square deviation (r.m.s.d.) of 1.72 Å was observed across a set of 10 proteins.
• The developed method demonstrated comparable accuracy to existing side chain building techniques.

Conclusions:

• The statistical atomic contact-based function effectively predicts protein side chain conformations.
• The method's ability to generate a limited set of likely conformations facilitates efficient combinatorial searches for multiple side chain predictions.
• This approach offers a valuable tool for protein structure refinement and computational protein design.