Re: RATTLE error with lipid bilayer system

From: Vermaas, Joshua (Joshua.Vermaas_at_nrel.gov)
Date: Tue Nov 21 2017 - 14:22:05 CST

Hi Ahmad,

Did any lipid atom numbers change? The restraints are based on the
atomic index if I remember correctly, so I am not surprised that when
you add components to your system (which may have changed the indicies)
that your simulation breaks if you use restraints based on the old
numbers, as you might be constraining things across molecules without
meaning too. The restraints are there to enforce chirality and cis-trans
isomerization of the lipids during minimization. If I were you, I'd
start from the minimized structure without the added components, add
your components in, and minimize/simulate without restraints.

-Josh

On 11/21/2017 12:12 PM, Ahmad Kiani Karanji wrote:
> Hello NAMD users
>
> I'm running POPC-POPG lipid bilayer system with hexamer blue
> proteorhodopsin inside it. I'm following the Charmm-gui input files.
> The raw files are working fine but I modified the system. I added some
> lipids in the middle of the protein and also retinal (resname RET is
> connected to the Lys 213) to the protein.
> My plan was to start with minimization and then going for
> equilibration but when the minimization finished my simulation
> crashed. It didn't even start equilibration.
>
> The error that I got is:
> ERROR: Constraint failure in RATTLE algorithm for atom 73327!
> ERROR: Constraint failure; simulation has become unstable.
> ERROR: Exiting prematurely; see error messages above.
> (atom 73327 is in lipid part of the system.)
>
> I also have this warning when I'm minimizing my structure:
> Warning: Bad global dihedral count!
> Warning: Bad global improper count!
>
> When I comment out the restraints (including planar restraint
> and dihedral restraint) simulation goes well, otherwise I get error.
> I running the system with 1fs timestep. I have no control on
> boundaries because charmm-gui define them by a formula. Different
> numbers of lipids doesn't work also. I tried to increase minimization
> steps, this one doesn't work also.
> I think this problem doesn't have anything to do with my protein
> because I ran it with detergents successfully.
> It would be appreciated if someone give a suggestion to overcome this
> error.
>
> Thanks
> Ahmad

This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:47 CST