Predicting DNA-binding protein and coronavirus protein flexibility using protein dihedral angle and sequence feature

Area of Science:

• Biochemistry and Molecular Biology
• Structural Biology
• Computational Biology

Background:

• Protein flexibility is crucial for biological functions like molecular recognition and catalysis.
• Understanding protein dynamics is key to deciphering protein function.
• Experimental and computational prediction of flexibility in DNA-binding and Coronavirus proteins is challenging.

Purpose of the Study:

• To develop an accurate method for predicting protein flexibility, specifically for DNA-binding and Coronavirus proteins.
• To leverage protein dihedral angle information for enhanced flexibility prediction.
• To analyze the correlation between amino acid properties, dihedral angles, and residue flexibility.

Main Methods:

• Introduction of DihProFle, a novel method incorporating dihedral angle, evolutionary, and physicochemical properties.
• Prediction of protein flexibility for DNA-binding and Coronavirus proteins.
• Assignment of flexibility classes to individual protein sequence positions.

Main Results:

• DihProFle achieved improved prediction accuracy of 2.2% (non-strict) and 3.1% (strict) compared to methods using only evolutionary or physicochemical data.
• The method demonstrated superior performance over existing protein flexibility analysis techniques.
• Charged hydrophilic residues were found to be more prevalent in flexible regions.
• Specific protein dihedral angle ranges (e.g., ψ angle 91°-120°) were associated with rigid regions.

Conclusions:

• Protein dihedral angle information and physicochemical properties, particularly hydrophilic residues, are significant predictors of protein flexibility.
• DihProFle offers a more accurate approach to predicting protein flexibility, aiding in the study of critical protein families.
• The findings provide insights into the structural determinants of protein flexibility.