Speaker
Prof.
Thomas Hamelryck
(University of Copenhagen)
Description
The so-called protein folding problem is the loose
denominator for an amalgam of closely related, unsolved
problems that include protein structure prediction, protein
design and the simulation of the protein folding process. We
adopt a unique probabilistic approach to modeling
bio-molecular structure, based on graphical models and
directional statistics [1,2,3,4,5]. Notably, we developed
the first probabilistic model of protein structure in full
atomic detail [1, 4]. In this talk, I will give an overview
of how rigorous probabilistic models of something as
complicated as a protein structure can be formulated,
focusing on the use of graphical models and directional
statistics to model angular degrees of freedom. I will also
discuss the reference ratio method [6], a novel statistical
method that can be seen as a surprising Bayesian variant of
the maximum entropy method. This method also sheds an
entirely new light on the in protein structure prediction
widely used potentials of mean force, which were up to now
poorly understood and justified. Finally, I will describe
some applications, including the investigation of protein
dynamics and the statistical inference of protein structure
from nuclear magnetic resonance (NMR) data [7] and small
angle X-ray scattering (SAXS) data [8].
[1] Hamelryck, T., Kent, J., Krogh, A. (2006) Sampling
realistic protein conformations using local structural bias.
PLoS Comput. Biol., 2(9): e131.
[2] W. Boomsma, K.V. Mardia, C.C. Taylor, J.
Ferkinghoff-Borg, A. Krogh, and T. Hamelryck. (2008) A
generative, probabilistic model of local protein structure.
Proc. Natl. Acad. Sci. U S A, 105(26):8932–8937.
[3] Frellsen, J., Moltke, I., Thiim, M., Mardia, KV.,
Ferkinghoff-Borg, J., Hamelryck, T. (2009) A probabilistic
model of RNA conformational space. PLoS Computational
Biology, 5(6), e1000406.
[4] Tim Harder, Wouter Boomsma, Martin Paluszewski, Jes
Frellsen, Kristoffer Johansson, and Thomas Hamelryck. (2010)
Beyond rotamers: a generative, probabilistic model of side
chains in proteins. BMC Bioinformatics, 11(1):306.
[5] M. Paluszewski and T. Hamelryck. (2010) Mocapy++ – a
toolkit for inference and learning in dynamic bayesian
networks. BMC bioinformatics, 11(1):126.
[6] Thomas Hamelryck, Mikael Borg, Martin Paluszewski, Jonas
Paulsen, Jes Frellsen, Christian Andreetta, Wouter Boomsma,
Sandro Bottaro, and Jesper Ferkinghoff-Borg (2010).
Potentials of mean force for protein structure prediction
vindicated, formalized and generalized. PLoS ONE, 5(11):e13714.
[7] Simon Olsson, Wouter Boomsma, Jes Frellsen, Sandro
Bottaro, Tim Harder, Jesper Ferkinghoff-Borg, and Thomas
Hamelryck. (2011) Generative probabilistic models extend the
scope of inferential structure determination. J. Magn.
Reson., 213(1):182-6.
[8] Stovgaard, K., Andreetta, C., Ferkinghoff-Borg, J.,
Hamelryck, T. (2010) Calculation of accurate small angle
X-ray scattering curves from coarse-grained protein models.
BMC Bioinformatics, 11:429.
