RE: Neutralize a box for FEP

From: Radak, Brian K (
Date: Wed Apr 12 2017 - 11:14:26 CDT


NAMD performs rigorous PME at both endpoints - all calculations involving A and B will be conducted without any knowledge of the other (yes, this means there are two energy/force computations per step).

The alchDecouple keyword has puzzling syntax IMO. "alchDecouple off" means that _all_ nonbonded interactions with each alchemical group are scaled by the appropriate lambda factor. "alchDecouple on" means that non-bonded interactions _between alchemical atoms of the same group_ are _never_ scaled. The latter requires additional PME calculations (up to 5 total I think?), but these are very small systems with low cost. This might exacerbate PME artifacts for charged systems, since you have to introduce additional neutralizing plasma terms, but I've never used "alchDecouple on", so I have no data to base that assertion on.


Brian Radak
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory

9700 South Cass Avenue, Bldg. 240
Argonne, IL 60439-4854
(630) 252-8643

From: [] on behalf of David Huggins []
Sent: Wednesday, April 12, 2017 10:30 AM
Subject: namd-l: Neutralize a box for FEP

Dear All,

I am interested in the correct way to neutralize a box in which a
charged molecule A is mutated by FEP into a molecule B with the same

Lets say A and B both have a charge of +1. Should counterions be added
to neutralize just one of the charges (-1 charge added) or both
charges (-2 charge added)?

My guess is that the former is correct, but it may be an issue that
the decoupled atoms interact with their periodic images. Does
alchDecouple then need to be set to off?




Dr David J. Huggins
University of Cambridge
Department of Chemistry
Lensfield Rd
United Kingdom

Phone: +44 (0)1223 763854


This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:20:12 CST