From: Markus K. Dahlgren (markus.dahlgren_at_yale.edu)
Date: Wed Sep 07 2011 - 12:54:16 CDT
Hi Tony,
It sounds like a forcefield parameter related problem. Either your starting
structure is wrong, or your parameters are. Make sure all residues are
connected properly with BONDS. Then look at your force field parameters and
your starting structure. The MD will adjust bond lengths according to your
parameters in your .prm file.
Best,
Markus
Quoting Anthony Rey <arey_at_etud.insa-toulouse.fr>:
> Hi,
>
> Thanks for accepting my email address on the mailing list. I have
> been using NAMD for a while now but can't figure out what's going on
> with the simulation.
> I am basically doing the same simulation as the Ubiquitin unfolding
> tutorial except I use it for a protein structure that I solved by NMR.
> I created the psf from the pdb and made sure that I didn't get any
> error messages (except from the N-terminal and C-terminal residues :
> poorly guessed residues).
>
> Then, I ran a script to put my protein into a water sphere. The
> output files look normal (both psf and pdb). I ran the simulation
> (5000 minimisation steps + 50 000 steps for the simulation), put the
> right diameter and center of the sphere for the spherical boundary
> conditions into the configuration file.
>
> When I look at the output, the minimization was long enough and the
> system looks quite stable at the end of the simulation. However, I do
> get some abnormalities when I look at the structure into VMD: the
> bonds get bigger during the simulation.
>
> The bonds are bigger but they are not huge. From what I've seen in
> the forum, a lot of people have similar issues except that they got
> huge bonds which I don't. I don't have long bonds at the beginning of
> the simulation, and I don't have any atoms that have {0.00 0.00 0.00}
> coordinates in the pdb either. The bonds just increase in length
> during the experiment, and the longer the experiment, the longer the
> bonds....
>
> Is is normal that the bond length increases with the simulation? Is
> there a way to fix this?
>
> I would really appreciate your help as I have no idea how to fix
> this. Please let me know if you need further details about the
> simulation.
>
> Best regards,
>
> Tony
>
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:57:43 CST