From: Alex Balaeff (abalaeff_at_polarisqb.com)
Date: Mon Dec 14 2020 - 20:13:27 CST
Dr. Alexander Balaeff
Polaris Quantum Biotech
On Mon, Dec 14, 2020 at 7:54 PM Marcelo C. R. Melo <melomcr_at_gmail.com> wrote:
> Yes, NAMD will take into account the neighboring images when selecting solvent molecules for electrostatic embedding. So if during the simulation, your molecule moves towards an edge, the water molecules on the other side of the box will be unwrapped to create the full electrostatic embedding. This will also happen if the QM region itself is "broken" by PBC wrapping (for example, if your QM region consists of two molecules bound by non-bonded interactions, and one gets moved to the other side of the box during the simulation), NAMD will first unwrap the QM region and then arrange solvent molecules so that they are as close to a QM atom as possible.
> On Mon, 14 Dec 2020 at 18:30, Alex Balaeff <abalaeff_at_polarisqb.com> wrote:
>> Dear NAMD developers and gurus:
>> Could those who know these details kindly advise how NAMD treats the
>> periodic images of point charges included in the QMMM calculations?
>> Say, I have a QM ligand simulated in an MM water box, and the ligand
>> drifts to the edge of the box. Is NAMD smart enough to include the
>> water molecules from the neighboring periodic cell into the MM shell
>> around the ligand that is sent to the QMMM calcs? Or is it not, and
>> one better make sure that the ligand is restrained to the center of
>> the cell?
>> If it matters, I am using a NAMD/MOPAC combo and run semiempirical QM
>> calcs (AM1, PM7).
>> Thank you all,
>> Dr. Alexander Balaeff
>> Polaris Quantum Biotech
This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:15 CST