[gmx-users] mutation to/from proline

[Mark Abraham]

On 25/04/2010 10:18 PM, afsaneh maleki wrote:
> Hi,
>
> I want to calculate relative free energy associated to mutation of P1
> (native protein) to p2 (mutated protein).In this mutation, Isolusine (in
> P1) is mutated to Proline(in P2). With using Thermodynamic cycle:
>
> Bilayer+P1= = => Bilayer-P1 delta G1 (association P1)
>
> Bilayer+P2= = => Bilayer-P2 delta G2 (association P2)
>
> Where P1 = = => P2 deltaG3
>
> Bilayer-P1 = = = > Bilayer-P2 deltaG4
>
> So, delta deltaG= (delta G1(association P1)- deltaG2 (association P2))=
> (deltaG3- deltaG4)
>
> To define the B state that will be used for the free energy calculation.
> For this; I need to specify the new atom type of each mutated atom and
> the set of parameters to it.
>
> But my question is that with regarding to special structure of proline,
> is it possible to update topology file? Who was done mutation of proline
> before? I searched but I didn’t find! Or is there manual for mutating
> from/to proline?

That's a pretty significant perturbation to be trying to study, because 
there have to be protein backbone rearrangements to accommodate proline, 
mutation of methyl to H, and a free alkyl chain losing its entropy and a 
pair of H. I expect you'll need to do a phenomenal amount of sampling to 
get useful results, even if you can think of a sensible mutation 
reaction coordinate. It would not surprise me if such a mutation had not 
been reported in the literature.

Accordingly, you might be best served by such methods implemented for 
implicit solvation in some other simulation package (if such exists).

Mark