| In this work we perform a statistical analysis of experimentally solved protein structures to study correlations between the properties of amino acid sequence and its final native structure, and address the following problems. (1) We study a phenomenon of conformational variability observed in the prion protein (PrP). A structural conversion of a normal, mostly α-helical cellular PrP into a pathogenic β-sheet-rich conformation causes fatal neurodegenerative diseases. First, we analyze the properties of structurally ambivalent sequence fragments (SASFs). Each SASF adopts two different conformations in two different proteins. Second, the data derived from the analysis of SASFs are used to study the properties of the prion protein, and gain insights into possible mechanisms of structural conversion. We show that the prion protein contains at least two sequence fragments with highly unusual intrinsic propensities, PrP(114–125) and helix B. These fragments are the primary candidates for conversion into β-sheet conformation and potential oligomerization sites. We also show that most PrP mutations associated with neurodegenerative disorders significantly increase the degree of local conformational variability. (2) We present a novel method designed to analyze and optimize the discriminative ability of knowledge-based potentials of mean force with respect to the 20 residue types. The method is based on the preference of amino acids for specific types of protein environment, and uses a virtual mutagenesis experiment to estimate how much information a given potential can provide about environments of each amino acid type. This method also allows one to obtain potential-specific matrices of distances between the amino acid types. The weighted combinations of these potential-specific matrices are optimized in order to obtain an amino acid exchange matrix which performs optimally in sequence alignment. (3) We perform a comparative analysis of correlations between global topological features of protein structure, sequence properties and folding rate in small single domain proteins with two-state kinetics. A novel correlation between the rate of folding and average intrinsic structural propensity of amino acid sequence is reported. |
