From: Vermaas, Joshua (Joshua.Vermaas_at_nrel.gov)
Date: Fri Aug 12 2016 - 14:45:00 CDT

Hi Jack,

Usually, the SMD forces should be high enough to force movement, but not high enough that everything else stays still by comparison, especially over several timesteps. Odds are you are driving your system far too hard, and crashing as a result. What does your NAMD configuration file look like?

-Josh

On 08/12/2016 01:31 PM, Jackson Cavett wrote:
Hello All,

I have been trying to equilibrate and minimize integrin alphaVbeta3 in a waterbox. I finished up a long equilibration simulation (about 40 ns), but am running into the "ERROR: Constraint failure in RATTLE algorithm for atom 994!" when I try to run a constant-force simulation. After reading up on this error, I changed the dcd frequency to every timestep and looked at the dcd file in VMD. When looking at the dcd file, it looks as though the smd atoms being pulled (by about 700 pN) move while every other atom stays fixed. It seems to me (though I am far from an expert) that the energy has not reached a constant value and therefore I would need more minimization steps. Before the equilibration, I minimized the system with three different minimzation simulations which were 10,000 timesteps each. My system has about 170,000 atoms, which is very large. Would anyone know how many minimization steps this system would need? Also, I could be on the wrong track with what my problem is, so if anyone has any other ideas t
hat would be great. Thank you all for your help!

Thanks,

Jack