lunduniversity.lu.se

Denna sida på svenska This page in English

Söderhjelm research group

Group members

  • Pär Söderhjelm (Senior lecturer)
  • Soumendranath Bhakat (PhD student)
  • .... Do you want to join our group? See postdoc notice below. We also have plenty of ideas for interesting diploma work projects.

Former group members

  • Eric Fagerberg (Master student, 2016)
  • Emil Åberg (Master student, 2016-2017)
  • Johanna Schmidt (project student, 2015)
  • Nils Hermansson Truedsson (project student, 2014)
  • Teodor Rodin (project student, 2014)

Postdoc position available

We are recruiting a fully financed postdoctoral researcher who should start in September 2018 or earlier. The research focus will be on method development in computational modeling of protein dynamics, involving one or several of the following areas:

  • Development of enhanced-sampling methods, especially strategies for automated selection of collective variables for metadynamics.
  • Testing and development of polarizable force fields for proteins.
  • Development of better strategies for validating computational results against experiments.

The exact research focus can be adapted to the postdoctoral researcher's own capabilities and interests so it is important to contact Pär Söderhjelm at an early stage to discuss possible projects. A detailed proposal is available here.

Molecular dynamics of proteins

Molecular dynamics simulations offer a fantastic opportunity to see how proteins work. By following the movement of every single atom under the influence of the forces from all other atoms, we get an extremely detailed picture of protein dynamics. Thus, we can nowadays do discoveries in the supercomputer instead of in the lab. 

 

However, there are two problems:

  • The time interval that one can simulate (typically a microsecond) is too short for directly watching biologically interesting processes, such as conformational changes, ligand binding, and enzymatic reactions. 
  • The force fields used to model the interactions between atoms are not sufficiently accurate to always give reliable results.

We try to solve both these problems by developing new simulation algorithms and interaction models, as well as adapting existing methods to new systems. We validate the theoretical results by comparing to state-of-the-art dynamical NMR experiments. Our long-term goal is to understand proteins and the various functions they perform.

Ongoing projects


Improved sampling

Improved Potentials

  • Solvation free energies using polarizable force fields
  • PMISP: a method for computing protein-ligand interaction energies at a high QM level
  • QM/MM methods with polarizable embedding

Selected publications

Sampling

Potentials

Collaborations