Re: [AMBER] MBAR input keywords not recognized in sander.MPI (Amber24)

From: Timothy Giese via AMBER <amber.ambermd.org>
Date: Mon, 20 Oct 2025 14:34:28 +0000

The valid options for sander are:
ifmbar ! turn on/off
bar_intervall ! output frequency
bar_l_min ! the minimum lambda (often 0.0)
bar_l_max ! the maximum lambda (often 1.0)
bar_l_incr ! the increment between states

Unlike pmemd, one currently cannot input arbitrary schedules — one is limited to uniform schedules.

If you are trying to perform alchemical free energy (AFE) transformations involving softcore functions, I highly recommend using pmemd.cuda (not sander and not the cpu version of pmemd). All of the AFE developments in the past 5 years have been made to the pmemd cuda code, and the pmemd cpu version was never updated to reflect the enormous number of improvements, bug fixes, and additions. One would typically only use sander to calculate "bookend" corrections via the indirect method (transformations that change potential energy functions rather than the chemical identity of a molecule). For example, the free energy of replacing a MM potential with a QM/MM potential.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Oct 20 2025 - 08:00:02 PDT
Custom Search