Relationships between amino acid sequence and backbone torsion angle preferences

Area of Science:

• Computational Biology
• Protein Structure Analysis
• Biophysics

Background:

• Understanding protein backbone torsion angles is crucial for predicting protein structure and function.
• The Rotational Isomeric States (RIS) model is a key tool for chain statistics in biophysical studies.
• Correlations between torsion angles can significantly impact protein folding pathways and stability.

Purpose of the Study:

• To calculate statistical averages and correlations of backbone torsion angles for chymotrypsin inhibitor 2.
• To investigate the influence of protein sequence context on torsion angle preferences.
• To compare statistical weights derived from knowledge-based energy maps versus excluded volume models.

Main Methods:

• Application of the Rotational Isomeric States (RIS) model for chain statistics.
• Derivation of statistical weights for phi-psi (torsion angle) pairs using Protein Data Bank (PDB) derived energy maps.
• Generation of random coil configurations using an all-atom protein model with volume exclusion for comparison.

Main Results:

• Protein Data Bank (PDB) data reveals significant correlations between adjacent backbone torsion angle pairs.
• These sequence-dependent correlations favor the selection of native-state torsion angles.
• Excluded volume effects alone do not produce context-dependent correlations influencing native-state torsion angle selection.

Conclusions:

• Protein sequence context plays a critical role in determining backbone torsion angle preferences.
• Knowledge-based approaches accurately capture sequence-specific correlations favoring native protein conformations.
• Simple steric considerations are insufficient to explain the observed context-dependent torsion angle preferences.