Related Experiment Videos
New efficient statistical sequence-dependent structure prediction of short to medium-sized protein loops based on an
J Wojcik1, J P Mornon, J Chomilier
1Systèmes Moléculaires et Biologie Structurale Laboratoire de Minéralogie-Cristallographie (LMCP), Universités Paris VI et Paris VII, Cedex 05, Paris, CNRS UMR7590, France.
Journal of Molecular Biology
|June 22, 1999
Summary
This study developed an automatic loop modeling algorithm for proteins. The algorithm accurately predicts protein loop structures, improving protein modeling accuracy and efficiency.
Area of Science:
- Structural bioinformatics
- Computational biology
- Protein structure prediction
Background:
- Protein loops are crucial for protein function and structure.
- Accurate modeling of loops remains a challenge in structural bioinformatics.
Purpose of the Study:
- To develop and validate an automatic algorithm for modeling protein loops.
- To improve the accuracy and efficiency of protein loop modeling.
Main Methods:
- Derived a database of 13,563 protein loops (3-8 amino acids).
- Performed statistical analyses of loop conformations and residue occurrences.
- Developed a loop modeling algorithm based on sequence-structure correlations and clustering.
- Validated the algorithm using a Jackknife test and compared it to CASP3 predictions.
Main Results:
- Identified length-dependent over-representations of specific amino acids and conformations in loops.
- Clustered loops into families based on backbone structure, revealing distinct sequence and conformational properties.
- Achieved average root-mean-square deviation (rmsd) predictions ranging from 1.1 Å (3-residue loops) to 3.8 Å (8-residue loops).
- Developed a statistical reliability score to enhance prediction accuracy.
Conclusions:
- The developed algorithm provides accurate and robust protein loop modeling.
- The statistical reliability score improves model selection accuracy.
- This tool is valuable for practical protein modeling, even with imprecise loop delimitation.