Clustering Techniques for Protein Fragments Analysis

Protein structure, Protein Data Bank, clustering, density-based clustering, sequence-structure relationships in proteins, clustering of protein fragments, protein structure prediction, sequence-structures fragments search.

This book presents a novel method for clustering of protein substructures that we developed in order to study the relationships between protein sequences and their structure. We show some properties of this method and the experimental results obtained by analyzing real data from PDB (Protein Data Bank). Furthermore, we present the results of the comparison of our method and the most commonly used methods for clustering of protein structures. The main advantage of our method is its high efficiency and scalability, which are key factors for analyzing large data sets. Finally, we propose a procedure for finding sequence profiles that tend to occur in more than one structural conformation but the number of their structural conformations is limited. This procedure is based on our method for protein substructure clustering.