[gmx-users] Change in structure after solvation_Gromacs

[ms]

On 05/12/10 13:20, Justin A. Lemkul wrote:
>
>
> ms wrote:
>> On 05/12/10 10:39, swagata chakraborty wrote:
>>> I was trying to run MD simulation of my protein in DMSO. I used the
>>> force
>>> field ffG53a6 and did the energy minimization after solvating the
>>> protein in
>>> DMSO box.
>>> But after solvating in DMSO, the structure is changing. My protein is a
>>> homodimer and the loop at the dimer interface is converting into a
>>> sheet.
>>> Please suggest where I am going wrong since the structure shouldnot
>>> change
>>> after solvation.
>>
>> You're not giving nearly enough detail (what is the dmso/water ratio?
>> what is your .mdp file?) etc.
>>
>> In any case, it wouldn't be the first time that simulations and
>> experiments disagree. How do you *know* that structure shouldn't
>> change? Have you tried using another force field, like OPLS?
>>
>
> This might just be a visualization artifact, something that gets posted
> here routinely. Some slight structural changes may have occurred during
> EM, such that the visualization program guess sheet instead of loop.
> 53A6 does tend to favor extended structures, but I would only suggest
> something is wrong if (1) a real simulation is conducted and the
> spurious secondary structure persists and (2) real analysis is done
> (like DSSP or STRIDE) to conclude the secondary structure, rather than
> just looking at it, as all visualization software uses different
> heuristics to assign structures. It does not make for convincing evidence.
>

You are absolutely right (as almost always you are!), yet more detail, 
the better. I'd say that visualizing it in Pymol using the dss command 
to give the DSSP assignment could help.

-- 
Massimo Sandal, Ph.D.
http://devicerandom.org