[AMBER] simulations of enzyme with ligand

From: Qinghua Liao <scorpio.liao.gmail.com>
Date: Thu, 16 Nov 2017 17:51:19 +0100


I did some simulations of an enzyme in complex with a ligand, the
complex sits in a octahedral box with water molecules.
I applied some distance restraints between the ligand and the enzyme in
order to keep the ligand stable at the binding site,
the force constant was set as 5 kcal/mol A2.

However, I found that there are some snapshots that the ligand goes out
of the binding site. I thought that it could be
the PBC problem, but I still got the long distance between the ligand
and the enzyme after I did the correction with autoimage.
I then check the output file, I found the restraints energy was
extremely high than the other frames.

Here are the distances calculated using cpptraj with autoimage
correction (only one of four is present):
# frames     distance (angstrom)
    14999       3.1086
    15000       2.8088
    15001      52.0178
    15002      14.3344
    15003       3.4264
    15004       3.0565

And here are the restraints from the output file:
# frames    Restraints (kcal/mol)
    14999         0.0350
    15000         0.4185
    15001      1209.4373
    15002       294.3320
    15003       20.4252
    15004       20.2892

As I did the production runs step by step, each step is for 50 ns (5000
frames). It means that the jumping-out happened
at the first frame of step 4 (frame 15001). I then checked the last
frame of step 3 ( frame 15000) from the restart file,
I found that the ligand was out of the binding site, and the restart
file which was not done with PBC correction (the ligand is still out of the
enzyme) was used for the the next step simulation. I did the simulations
with iwrap=1 and ioutfm=1.

Any comment and suggestion? Thanks a lot!

All the best,

AMBER mailing list
Received on Thu Nov 16 2017 - 09:00:06 PST
Custom Search