Team:EPF-Lausanne/Modeling overview



Modeling overview

Protein domain of interest

Our protein of interest is LOVTAP. This protein was synthetically engineered by Pr. Sosnick's group from the University of Chicago. It is a fusion protein between a LOV domain (Avena Sativa phototropin 1) and the E. Coli tryptophan repressor. This protein undergoes changes under light activation as shown by Sosnick et al, namely when the protein is activated by light it binds to DNA and inversely.

For more information about LOVTAP protein please click here.


Sosnick et al. found that light-activated LOVTAP isn't stable. After light excitation, the LOV domain returns to its ground state (non light-activated state) very quickly.

So the aim of the molecular dynamics simulation is to simulate the LOV domain in its environment under light activation (so-called light state) and without light activation (ground state, so-called dark state), calculate atom and residue movements of particular/interesting LOV domain regions, and finally deduce which residue(s) could be mutated to stabilize the light-activated state of this LOV domain (increase its lifetime).

Then, simulation of the complete LOVTAP protein with selected mutations could give us insights about the behaviour of our protein in its environment.

Starting material

Both LOV domain crystallography files were obtained from RCSB:

Light-activated LOV domain
Dark LOV domain

These crystallographies were done by Halavaty et al..

Molecular dynamics: a little theory

Molecular dynamics simulation consists in the numerical, step-by-step, solution of the classical equations of motion. For this purpose we need to be able to calculate the forces acting on the atoms, and these are usually derived from a potential energy.

         Click here to expand


The following information is mostly taken from an Introduction to Molecular Dynamics: see here their web page.

1. Minimization

Using the forcefield that has been assigned to the atoms in the system, it is essential to find a stable point or a minimum on the potential energy surface in order to begin dynamics. At a minimum on the potential energy surface, the net force on each atom vanishes.

         Click here to expand

2. Equilibration

Molecular dynamics solves the equations of motion for a system of atoms. The solution for the equations of motion of a molecule represents the time evolution of the molecular motions, the trajectory. Depending on the temperature at which a simulation is run, molecular dynamics allows barrier crossing and exploration of multiple configurations.

         Click here to expand

3. Simulation

We run a 100ns simulation, from which we will collect the data and see what happens to our protein! We made calculations during nearly 4 weeks, on 64 processors. The computational part of the project was supported by the HPC infrastructures at LBM: the available resource was constituted by a local Linux Cluster of 200 AMD Barcelona cores which we were be granted access at various phases of the project.

It was made for both light and dark states, the topography and parameters files used were taken from the article from Schulten , Dynamic Switching Mechanisms in LOV1 and LOV2 Domains of Plant Phototropins (see here for more details ont he reference).

4. Simulation Analysis

This part is dedicated to the analysis of simulation results. The goal is to find some interesting clues to understand the conformational change of our protein upon light activation.

We analysed the atom movement using calculations of dihedral angles, atom distances and H bond distances in order to find the initiation of the general conformational change of our protein upon light activation.

PCA analysis (Principal Component Analysis) would have been useful to make predictive models of the Jalpha helix movement, click here for more information.

For more details about what we have done, see :

  • the Analysis Methods page, which is composed of a step-by-step description of what we did : click here for more information on this topic.
  • the Results page, which explain what we elicited from our raw data: click here for more information.

Back to top