From: Christian Seitz (
Date: Mon Jan 06 2020 - 18:37:42 CST


I have been running a long NAMD simulation of a homotetramer protein in 12
chunks. A few months ago I finished the first 6 chunks; one monomer of the
tetramer would flip to a different box during some chunks. So I combined
those trajectories into a single trajectory and wrapped it using the TK
Console with:

pbc wrap -centersel "protein" -center com -compound residue -all
pbc unwrap -first now

Now, I've finished all 12 chunks of the simulation and again combined the
chunks of trajectories into a long one. But now, I cannot wrap the full
trajectory or any of the shorter chunks anymore, and the pbc boundaries get
deleted. For example, directly after loading a trajectory:

% set cell [pbc get -now]
{169.186295 169.186295 226.764709 90.000000 90.000000 90.000000}
% pbc wrap -centersel "protein" -center com -compound residue -all
ERROR) Suspicious pbc side length (a=0.000000 b=0.000000 c=0.000000). Have
you forgotten to set the pbc parameters?
domain error: argument not in valid range
% set cell [pbc get -now]
{0.000000 0.000000 0.000000 90.000000 90.000000 90.000000}

This happens regardless of the wrap commands I use. I can reinstate the
boundaries with pbc set, but I then cannot get wrapping to work properly
(it takes hours and in the end has some atoms in a different box, even
after unwrapping or using different wrap commands). I have scoured the VMD
mailing list and other sites, and haven't found anybody else with this
problem. Any help would be greatly appreciated. Thanks!

Best regards,

*Christian Seitz*
PhD Student, Biochemistry & Biophysics | UC-San Diego
NSF GRFP Fellow, Amgen Scholar
McCammon lab <> and Amaro lab
<> <>