Conformational Ensembles from Experimental Data and Computer Simulations

Conformational Ensembles from Experimental Data

and Computer Simulations

Poster Abstracts

34-POS

Board 34

Multiscale simulations of partially disordered systems: Representing environment-induced

helix-coil transitions

Christoph Globisch

, Cahit Dalgicdir

, Mehmet Sayar

, Christine Peter

University of Konstanz, Konstanz, Germany,

University of Darmstadt, Darmstadt, Germany,

Koç University, Istanbul, Turkey.

Coarse grained (CG) models are widely used to study peptide self-assembly and nanostructure

formation. One of the recurrent challenges in CG modeling is the problem of limited

transferability. A crucial question for peptides is whether a model reproduces the molecule's

conformational response to a change in its molecular environment. Examples are conformational

transitions between a rather disordered and an ordered state upon a change in pH value or due to

the presence of a soft apolar/polar interface. To handle such transitions CG models mostly utilize

auxiliary interactions to aid secondary structure formation. Such interactions take care of

properties of the real system that are per se lost in the coarse graining process such as dihedral-

angle correlations along the backbone or backbone hydrogen bonding. Since the CG models are

designed to emphasize certain conformational propensities they may destroy the ability of the

model to respond to stimuli and environment changes. This points out how important it is to

investigate whether they impede transferability.

To analyze such processes in combined atomistic/CG manner a common characterization of the

shallow conformational free energy landscapes is needed which is dominated by a huge number

of metastable and often ill-defined minima. Dimensionality reduction methods such as

multidimensional scaling-like embedding (sketch-map) can be applied to compare the phase

space sampled at both resolutions (atomistic/CG), either to judge the success of elevated

sampling techniques and possibly guide further simulations, or to monitor the response of the

systems to external stimuli.