HCPM--program for hierarchical clustering of protein models.

Area of Science:

• Computational biology
• Structural bioinformatics
• Protein structure prediction

Background:

• Accurate clustering of protein structures is crucial for analyzing results from comparative modeling and ab initio structure prediction.
• Existing methods may not efficiently handle the diverse structural data generated by various prediction techniques.

Purpose of the Study:

• To introduce Hierarchical Clustering of Proteins (HCPM), a novel tool for grouping protein structures.
• To develop and validate an effective clustering algorithm for protein structure datasets.

Main Methods:

• Implementation of a hierarchical clustering algorithm tailored for protein structure data.
• Development of a heuristic approach for optimal cluster identification.
• Testing and validation of the HCPM tool within the context of the CASP6 experiment.

Main Results:

• The hierarchical clustering algorithm effectively groups protein structures from comparative and ab initio modeling.
• The heuristic provides a reliable method for selecting optimal clusters.
• Successful application of HCPM in the CASP6 experiment demonstrates its practical utility.

Conclusions:

• HCPM offers a robust solution for clustering protein structures derived from computational methods.
• The developed algorithm and heuristic enhance the analysis of large-scale protein structure prediction datasets.
• Validation in CASP6 confirms HCPM's capability in real-world structural bioinformatics challenges.