Related Experiment Videos
Effects of limited input distance constraints upon the distance geometry algorithm.
C M Oshiro1, J Thomason, I D Kuntz
1IBM Palo Alto Scientific Center, California 94304.
Biopolymers
|August 1, 1991
Summary
This study analyzes the distance geometry (DG) algorithm for protein structure determination. Findings reveal DG structures are more compact than actual ones and require sufficient distance constraints for accuracy.
Area of Science:
- Computational Biology
- Structural Biology
- Bioinformatics
Background:
- The distance geometry (DG) algorithm is crucial for determining protein structures from experimental data.
- Understanding the algorithm's parameters, such as distance bounds and constraint numbers, is essential for accurate structure prediction.
Purpose of the Study:
- To critically evaluate the distance geometry (DG) algorithm used in protein structure determination.
- To investigate the impact of bound smoothing, random distance selection, and the number of constraints on structural accuracy.
Main Methods:
- Computational experiments using simulated and real data for basic pancreatic trypsin inhibitor (BPTI).
- Analysis of NMR and crystallographic measurements to validate distance geometry predictions.
- Development of a model to describe the behavior of randomly selected trial distances.
Main Results:
- Upper bounds from bound smoothing showed a linear relationship with true crystal distances.
- Proposed model explains results from randomly selected trial distances.
- DG-generated BPTI structures were found to be more compact than the native crystal structure.
Conclusions:
- Protein backbone resemblance to test structures diminishes when distance constraints fall below degrees of freedom.
- The number and selection of interresidue distance constraints significantly impact DG algorithm accuracy.
- Conclusions are likely generalizable to other versions of the DG algorithm, despite sensitivity to distance selection methods.