On Wed, Dec 10, 2014 at 5:28 AM, de Manzanos Guinot, Angela <
angela.de-manzanos11.imperial.ac.uk> wrote:
> Hi Jason,
>
> Thank you very much, I have written a script that does it and I obtain
> very similar results than the ones I obtained previously with MMGBSA.py,
> only from the third decimals are a bit different which looks like rounding
> issues, but nothing significant (as the energy terms of the receptor and
> comples are arounf exp5). If I set igb5, the saltcon and surften
> appropriate for my system, extdiel=80.0 and gbsa=1 it calculates the ESURF
> using LCPO too, so I don't need an extra calculation with cpptraj. Also,
> regarding imin=5 it doesn't work as it gives an error regarding the
> .inpcrd, looking at the amber manual I found this:
>
> IMPORTANT CAVEAT: The initial coordinates input file used (-c <inpcrd>)
> should be the
> same as the initial coordinates input file used to generate the original
> trajectory. This is because
> sander sets up parameters for PME from the box coordinates in the initial
> coordinates input file
>
> And as I generate the individual trajectories by stripping things from the
> raw trajectory, I cannot obtain the initial inpcrd of these, I tried all
> sort of things here.
​The above warning does not apply if you are using GB or any other
non-periodic system. If you are using GB, you can provide *any* compatible
inpcrd file (MMPBSA.py generates a restart file from the first frame of the
first trajectory and just uses that) -- all of the information from the
inpcrd file is ignored for GB calculations​.
imin=5 is actually quite a bit faster than doing a single-point energy
calculation using individual restarts. The reason is that when you use
imin=5, the topology file only has to be read once, and all of the data
structures that sander uses only have to be allocated and initialized
once. When you're doing a single-point energy calculation, the cost of
initialization is sometimes similar to the cost of a single energy
evaluation. And then of course imin=5 requires only 2 files to run a
calculation (the inpcrd and trajectory) -- using restarts requires as many
files as you have frames in the trajectory.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Dec 10 2014 - 04:30:02 PST
