Show simple item record

dc.contributor.advisorElber, Ron
dc.creatorChen, Szu-Hua
dc.date.accessioned2016-10-18T18:30:44Z
dc.date.available2016-10-18T18:30:44Z
dc.date.issued2016-08
dc.date.submittedAugust 2016
dc.identifierdoi:10.15781/T25H7BW1Q
dc.identifier.urihttp://hdl.handle.net/2152/41710
dc.description.abstractProteins are biological macromolecules that are involved in a wide range of cellular processes. The diverse functions of proteins are closely related to their dynamics and structures. Structures are frequently coded in a complex manner in the amino acid sequences. In this dissertation I discuss the dynamics of a special class of proteins through studies of their sequences and structures. These proteins are “switches,” which are made of highly similar sequences that fold to dramatically different structures. The existence of protein switches provides a great challenge to structure prediction algorithms as well as to our understanding of the process of protein structure evolution. To identify protein switches, we developed methods that assign switch sequences to structures with high accuracy. One method uses short MD simulations to enrich structural ensembles of protein switches in the neighborhood of their initial conformations for scoring by contact maps. The other method uses evolutionary profiles and contact maps of the wild-type proteins. Both methods were first tested against a series of experimentally engineered proteins in a switching system and then applied to examine a large number of computationally sampled protein switches for a particular pair of structures in sequence space. From the sampled switch sequences we found that making a point mutation near the N- and C-termini of the sequences is more likely to make the proteins switch between structures. To study the conformational change of a protein switch with a fixed sequence between two metastable states in conformational space, we proposed a new algorithm, named “Chain Growth”, to calculate reaction pathways. Unlike commonly used methods that require an initial guess of a path and minimize the energy of the path by local quenching, our method propagates the path in small segments and optimizes the whole path globally. These features avoid the problems of generating very distorted initial structures that other methods frequently encounter and allow more efficient minimization of the path. We provided computational examples of using Chain Growth to calculate the minimum energy path on the Müller potential energy surface as well as to the studies of conformational changes of alanine dipeptide and folding of tryptophan zipper.
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.subjectProtein switch
dc.subjectSwitch sequence sampling
dc.subjectReaction pathway
dc.subjectChain growth
dc.subjectMD simulation
dc.subjectContact energy
dc.subjectSecondary structure prediction
dc.subjectPoint mutation
dc.subjectBinary flip
dc.subjectHausdorff distance
dc.subjectDiscrete Fréchet distance
dc.titleProtein dynamics in sequence and conformational spaces
dc.typeThesis
dc.date.updated2016-10-18T18:30:44Z
dc.contributor.committeeMemberRen, Pengyu
dc.contributor.committeeMemberJohnson, Kenneth A.
dc.contributor.committeeMemberEllington, Andrew
dc.contributor.committeeMemberMakarov, Dmitrii E.
dc.description.departmentBiochemistry
thesis.degree.departmentBiochemistry
thesis.degree.disciplineBiochemistry
thesis.degree.grantorThe University of Texas at Austin
thesis.degree.levelDoctoral
thesis.degree.nameDoctor of Philosophy
dc.type.materialtext


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record