Team:EPF-Lausanne/Results/Differential Analysis

   

 

 

 

Differential analysis

The two state of the protein being well characterized by the 2 simulations, we focus now on finding any difference between the two.

=Movement of the alpha helix= As you can see in our simulations, we weren't able to see a clear movement of the helix. It would have been the nicest result possible, because it would have validated the hypothesis of the helix being a rigid lever. Anyway, as our simulations reproduce nicely the B-factor of the crystallography from the RMSF, we can deduce that they are valide. The lack of movement could have multiple origins:
 * it could result from a bad crystallography
 * more likely it could result from a poor charge description of the FMN.

We re-used the charge file of the supplementary data from Schulten's article, but we would need to do some quantum mechanic simulation to validate these, but we don't have the time in the perspective of the jamboree.

As said before, we do not see a clear change in the angle, that is, there is no clear movement of the helix.

=Sidechains involved in signal transmission from FMN to alpha helix= We learned from an article that the GLN513 seems to be involved in an hydrogen bond with FMN. We tried to monitor this bond in our dark light simulation without seeing anything, both GLN513 and ASN414 are quite stable.

So we tried to compute the RMSF of the sidechains to identify the most moving ones. If we can find one that has an hydrogen bond either with alpha helix or any residue near FMN, we would have a candidate for mutation. Here is the graph. The blue line is the mean RMSF value for each residue. Then the red line is the same value minus the RMSF of the CA. In this way, only the residue with moving sidechains have a high positive value. We can see that the CYS450 has an high value on this graph, which tends to prove that it has a signification.

So now, we want to see where are those residues located in the beta sheet. You can see the FMN in red and residues with high rmsf (resid 479 or resid 484 or resid 495 or resid 509 or resid 511) displayed in cpk non-red. Sadly, none of these residues are involved in an hydrogen bond.

We can deduce that maybe the charges of the FMN are not realistic.

=CYS450 and active site= The dark state simulation reproduce nicely the values of position for the CYS450. Furthermore we have similar repartition between ON and OFF orientation of the residue in the simulation and in the crystallography. The mutations we are going to choose will tend to change this ratio to make it more ON.

Here is a picture of the situation around CYS450 and FMN.