Protein katlanması dinamiğine optimal kontrol yaklaşımı

This thesis explores the pathways of protein folding using a 3-D off-lattice model. Adynamic model is proposed and used in an optimization framework. Amino acids arerepresented by monomer beads centered at alpha carbon atoms. The interactions betweenmonomers are represented through spring forces .In addition, a force field is introducedas a control input to the dynamic model. Next, protein folding is formulated as an optimalcontrol problem in which a particular form of energy is minimized subject to the dynamicmodel predictions and physical constraints. This approach allows us to generate possiblepathways of protein folding from an initial configuration to the given native state. Ourmodel is applied to a fast folding, 36-residue protein, villin headpiece sub domain. Thesimulated structures resemble the real native state of chicken villin headpiece with thealpha carbon based root mean square deviation of 3.97 angstrom on the average. Startingfrom several initial conditions that cover a wide range of compactness, the sequence ofdynamic events along trajectories are studied. Furthermore, the characteristics of forceson each bead and their relation to the constraints and interactions are analyzed.