Related Experiment Videos
Fold recognition and accurate sequence-structure alignment of sequences directing beta-sheet proteins.
Andrew V McDonnell1, Matthew Menke, Nathan Palmer
1Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA.
Proteins
|March 21, 2006
Summary
Predicting protein structure is vital for public health proteins like allergens. A new algorithm, BetaWrapPro, accurately identifies beta-helix and beta-trefoil protein structures, aiding in predicting their 3D forms.
Area of Science:
- Structural bioinformatics
- Computational biology
- Protein structure prediction
Background:
- Predicting protein structure from amino acid sequence is crucial for understanding protein function, especially for proteins of public health significance such as toxins, allergens, and cytokines.
- The parallel beta-helix and beta-trefoil protein folds are common structural motifs found in many functionally important proteins.
Purpose of the Study:
- To develop and evaluate a novel computational method for accurately predicting parallel beta-helix and beta-trefoil protein folds from sequence.
- To assess the performance of the developed algorithm against existing methods for motif recognition.
Main Methods:
- A new algorithm, BetaWrapPro, was developed, integrating pairwise beta-strand interaction probabilities with sequence profile evolutionary information.
- The algorithm incorporates a 'wrapping' component to model protein folding initiation and subsequent motif interaction.
- Performance was evaluated using cross-validation on a comprehensive database of known beta-helix and beta-trefoil structures from the Protein Data Bank (PDB).
Main Results:
- BetaWrapPro achieved high performance in recognizing beta-helix and beta-trefoil folds, with 100% sensitivity and 99.7% and 92.5% specificity, respectively.
- The algorithm accurately aligned a high percentage of residues (88% for beta-helices, 86% for beta-trefoils) to structural templates.
- Predicted structures were further refined using the SCWRL side-chain packing program, and a web server is available for public use.
Conclusions:
- BetaWrapPro significantly outperforms previous motif recognition programs for beta-helix and beta-trefoil folds.
- The method enables accurate prediction of protein structure, including the identification of unexpected structures like a parallel beta-helix in a pollen allergen.
- The availability of the BetaWrapPro web server facilitates the prediction of 3D structures for proteins with these important folds.