Protein folding is a crucial biological process in which long chains of amino acids acquire their specific three-dimensional structure, which is essential for their functionality. A highly simplified theoretical model of protein folding is the HP model. This model divides amino acids into hydrophobic (H) and polar (P) categories. The HP model is useful due to its simplicity and allows for the study of the basic principles of protein folding.
In this work, I first studied the structures of peptides composed of the amino acids leucine and serine using various computational methods with the Spartan program. I employed the molecular mechanics method MMFF, the semi-empirical method PM3, the Hartree-Fock method, and density functional theory. It turned out that Spartan is not suitable for studying the equilibrium geometries of peptides, as they contain too many atoms and are computationally too complex for Spartan to handle. Additionally, different computational methods yielded quite different results.
Then, I studied peptide folding using the HP model. For this purpose, I developed a program in Java that analyzes all unique conformations of peptides on a 2D square lattice. This program also allows for the analysis of somewhat longer amino acid sequences and can easily analyze HP sequences up to 20 amino acids in length. The HP model was not successful in predicting the structures of real proteins, but it is suitable for studying the general principles involved in peptide folding. Longer sequences with a sufficiently high proportion of hydrophobicity had more H-H contacts in the native structure (the most energetically favorable structure according to the HP model) and also had a smaller proportion of native structures. Thus, such sequences were more selective in folding.
|
