147 / 161
Conformational Ensembles from Experimental Data
and Computer Simulations
Poster Abstracts
143
106-POS
Board 26
Integration of SAXS Data into Biomolecular Simulations
Marie Weiel
, Ines Reinartz, Alexander Schug.
Karlsruhe Institute of Technology, Karlsruhe, Germany.
The cardinal aim of structural analyses in molecular biology and biophysics is to reveal an
interrelation between macromolecular structure and conformational changes on the one hand and
function of biological macromolecules on the other hand. Biological small angle X-ray scattering
(SAXS) is an experimental technique for structural characterization of macromolecules and
complementary to common high-resolution methods such as X-ray crystallography and NMR
spectroscopy. To date, SAXS data are often interpreted by ambiguous reconstruction of low-
resolution three-dimensional models from one-dimensional scattering intensities or assembly of
rigid high-resolution elements. However, with large structural rearrangements being involved,
these methods do not yield satisfying results. We include the limited information from SAXS
into molecular dynamics (MD) simulations using native structure-based models (SBM). A
particular initial structure, e.g. from X-ray or NMR methods, is defined as the global energetic
minimum in a minimally frustrated single-basin energy landscape dominated by native
interactions. The resulting description in terms of a smooth energy funnel can provide rich
information about complex processes and is computationally efficient. In order to incorporate
information from SAXS, a bias term is added to the SBM potential so that conformations
reproducing the experimental target data are energetically favoured. In this vein, SAXS data may
be reasonably interpreted whilst simultaneously retaining chemical knowledge and sampling
power of molecular force fields. Running SAXS-guided MD simulations of a protein in some
known initial configuration, one can obtain a well-grounded atomistic structure of the protein in
another conformation corresponding to the experimental data by dynamically fitting the starting
model to the SAXS intensity. Giving fast and reliable structure predictions for transiently
populated conformations and related conformational changes, we hope to make a significant
contribution to unraveling the relation between macromolecular structure and function.