A Closer Look on Protein Unfolding Simulations through Hierarchical Clustering

Abstract

Abstract-Understanding protein folding and unfolding mechanisms are a central problem in molecular biology. Data obtained from molecular dynamics unfolding simulations may provide valuable insights for a better understanding of these mechanisms. Here, we propose the application of an augmented version of hierarchical clustering analysis to detect clusters of aminoacid residues with similar bahavior. These clusters hold similar global pattern behavior of solvent accessible surface area (SASA) variation in unfolding simulations of the protein Transthyretin (TTR). Classical hierarchical clustering was applied to build a dendrogram based on the SASA variation of each aminoacid residue. The dendrogram was enriched with background information on the amino-acid residues, enabling the extraction of sub-clusters with well differentiated characteristics.