[AMBER] Thermodynamic integration equilibration

From: Gian Marco ELISI <gianmarco.elisi.unipr.it>
Date: Wed, 12 Jan 2022 11:23:01 +0000

Dear Amber users,

I have some questions about the equilibration performed prior to the thermodynamic integration. In most of the articles, I have found that a brief equilibration of the system is conducted with a λ=0.5 with both van der Waals and atomic charges of the TI regions controlled through soft core functions. However, I did not encounter any brief explanation as the references in the protocols just bounced me to older articles without any motif. Is there a specific reason for this approach?
If I try to start the equilibration from λ=0, I proceed with an incremental protocol where coordinates and velocities for a simulation with a x λ value are then adopted for the next simulation with λ value equal to x+1. This until the simulation crashes with λ=1. Furthermore, everything happens after λ=0.5 appears catastrophic with the ligand able to unbind and rebind in the next steps, events that do not happen in traditional MD. This happens with or without the partial charges of the TI region turned off, while attempts to restrain the system after the NVT phase are as well infructuous. These problems seem illogical as this setting of the simulation seems analogous to the one adopted in single step calculations.

Gian Marco

AMBER mailing list
Received on Wed Jan 12 2022 - 03:30:02 PST
Custom Search