Re: [AMBER] MD with emap restraints, restraints have no effect

From: David A Case <>
Date: Mon, 5 Jul 2021 08:11:50 -0400

On Mon, Jul 05, 2021, Will Gerrard wrote:

> EMAP = 87057040.6349

Is this value coming from a simulation with some high value of fcons?

EMAP was designed to refine structures that had already been fit into
electron density maps. (It comes from the days where only fairly
low-resolution maps were available, although seems to also work well at much
higher resolutions.)

With things like distance or RMS restraints, the force on the restrained
atoms increases as you get further away from the target. But with EMAP, you
reach a plateau in the energy as you get farther away -- if there is little
overlap between the current structure and the map, things don't change if
you move the structure around a bit via MD. This means that the force on
the restrained atoms goes to zero if the structure is not close to the map.

Such behavior can sometimes be valuable (see the ialtd parameter in NMR
restraints), but it doesn't seem to be what you want. If there is any
barrier between the starting and target structure (caused, say, by the
presence of the protein) you won't get the fitting you desire.

Would targetted MD meet your needs? That would seem to be a more
straightforward path.


AMBER mailing list
Received on Mon Jul 05 2021 - 05:30:03 PDT
Custom Search