Improved convergence and speed for the distance geometry program APA to determine protein structure

Area of Science:

• Biochemistry and structural biology
• Computational biology
• Bioinformatics

Background:

• The alternating projections algorithm (APA) is utilized for protein three-dimensional structure determination.
• Previous studies demonstrated APA's success with poly-L-alanine and reported its sampling and convergence properties using bovine pancreatic trypsin inhibitor (BPTI).
• A limitation identified was the slow asymptotic convergence of the APA.

Purpose of the Study:

• To enhance the convergence speed of the alternating projections algorithm for protein structure determination.
• To investigate the efficacy of a weighted quadratic penalty function in accelerating the final conformation generation process.

Main Methods:

• The study builds upon the established alternating projections algorithm (APA).
• A weighted quadratic penalty function was introduced to refine the conformational output after initial convergence by APA.
• Performance was evaluated by comparing the convergence time of the enhanced method against the standard APA.

Main Results:

• The alternating projections algorithm (APA) was confirmed to be successful in determining protein structures.
• The addition of a weighted quadratic penalty function significantly improved convergence speed.
• The penalty function achieved final conformation in half the time compared to relying solely on APA's asymptotic convergence.

Conclusions:

• A weighted quadratic penalty function offers a substantial improvement over the standard alternating projections algorithm for protein structure prediction.
• This enhanced approach accelerates the determination of accurate three-dimensional protein conformations.
• The findings suggest a more efficient computational strategy for structural biology.