Prediction of Long Loops with Embedded Secondary Structure using the Protein Local Optimization Program.

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Accurate protein homology modeling relies on precise prediction of protein loops, especially those with secondary structures.
• Excluding loops with secondary structures limits the scope of loop prediction methods.

Purpose of the Study:

• To extend the Protein Local Optimization Program (PLOP) for accurate prediction of protein loops containing helical or hairpin segments.
• To improve the accuracy and applicability of computational protein structure prediction.

Main Methods:

• Hierarchical conformational sampling and ranking using a molecular mechanics force field.
• Utilizing an alternative dihedral library for α-helical segments based on common helical (ϕ,ψ) angles.
• Applying the method to predict loops up to 17 residues (helical) and 13 residues (hairpin).

Main Results:

• Achieved sub-Ångström RMSD accuracy in 80% of helical loop predictions and 83% of hairpin loop predictions.
• Demonstrated restoration of perturbed loop conformations to native structures with sub-Ångström RMSD.
• Showcased high accuracy even with imprecise surrounding environments in homology modeling.

Conclusions:

• The enhanced PLOP method significantly improves the prediction of protein loops with secondary structures.
• This advancement broadens the utility of homology modeling for atomic-level protein structure determination.
• The method shows robustness in restoring native loop conformations under perturbed conditions.