A fast loop-closure algorithm to accelerate residue matching in computational enzyme design.

Area of Science:

• Chemical Biology
• Computational Chemistry
• Biocatalysis

Background:

• Designing functional active sites on inert scaffolds remains a significant challenge in chemical biology.
• Accurate placement and geometric constraints of catalytic residues are crucial for enzyme function.

Purpose of the Study:

• To enhance the enzyme design program ProdaMatch by integrating a novel Newton-direction-based fast loop-closure algorithm.
• To improve the accuracy and speed of catalytic residue matching for de novo enzyme design.

Main Methods:

• Incorporation of a Newton-direction-based fast loop-closure algorithm into the ProdaMatch program.
• Testing the algorithm's performance on loop-closure accuracy and iteration steps using a large dataset.
• Evaluating the revised ProdaMatch's ability to identify native active-site matches on various scaffolds.

Main Results:

• The novel loop-closure algorithm achieved high accuracy (99.51% within 0.05 Å) in under 400 iteration steps for over 64,000 loops.
• The majority of loops were closed within 100 iteration steps, demonstrating significant speed enhancement.
• The revised ProdaMatch successfully identified all native matches for ten scaffolds in active-site recapitulation tests.

Conclusions:

• The enhanced ProdaMatch program demonstrates high speed and accuracy in matching catalytic residues to scaffolds.
• This advancement facilitates scaffold selection in de novo enzyme design, potentially leading to enzymes with higher initial activities.
• The integration of complex theoretical enzyme models is now more feasible for improved enzyme design outcomes.