RE: SMD restraints

From: Allen, Caley R (
Date: Tue Nov 15 2016 - 13:30:17 CST

Oops. I will make sure to hit reply all.

Thank you.

I will do just that. And yes, PBC and wrap atoms has proven to be problematic.


Sent from my Verizon, Samsung Galaxy smartphone

-------- Original message --------
From: Giacomo Fiorin <>
Date: 11/15/16 2:18 PM (GMT-05:00)
To: "Allen, Caley R" <>
Cc: NAMD list <>
Subject: Re: namd-l: SMD restraints

Hi Caley, please respond to the mailing list.

The addgroup and addforce keywords are documented fairly clearly in my opinion, so I guess what you're asking is how to test whether the code will work as you expect.

I would recommend a simple test by computing the COM of those atoms in VMD (see the doc for the measure command there) and then printing the same COM with NAMD.

FYI, watch out for possible issues with PBCs and wrapping of atoms.


On Tue, Nov 15, 2016 at 2:04 PM, Allen, Caley R <<>> wrote:
Thank you Giacomo for your answer.

I have seen the websites before. If I may quickly ask, by defining a group and then use its COM and apply forces to, I would use the addgroup and list each atom number? So for instance the bilayer I would list atomids 129 -37061? I just want to make sure I understand .

Thanks again.
Best wishes.

Sent from my Verizon, Samsung Galaxy smartphone

-------- Original message --------
From: Giacomo Fiorin <<>>
Date: 11/14/16 4:18 PM (GMT-05:00)
To: NAMD list <<>>, "Allen, Caley R" <<>>
Subject: Re: namd-l: SMD restraints

On Mon, Nov 14, 2016 at 4:14 PM, Allen, Caley R <<>> wrote:
Greetings NAMD experts.

I was browsing through the mail list, in an effort to find a solution to a problem when I stumbled upon a post from Feb. 29, 2012. The interesting part (to me) reads as follows (Axel wrote the response in black):

What you say makes sense, but I want to compute the work done by the pushing
no. this is bad. you cannot change the force field.
force as I insert the peptide into the membrane (steered MD). Hopefully I'll
get some free energy profile as a function of insertion depth. Hence I need
the membrane to stay in place.
not quite. what you need is to have the steering force
being applied to both, the center of mass of the *entire*
membrane and the center of mass of the *entire* protein.
restraining a few lipid molecules in space will not help.
best case, it will artificially bend the membrane, worst
case, it will just rip those out.
in general, you should not worry too much about a little
drift, after all you have periodic boundary conditions,
and for a clean steered MD PMF on inserting a protein
into a membrane, you need quite a bit of water in between.

I am attempting to do something very similar – I would like to “pull” a ligand off of the bilayer leaflet via SMD at a constant velocity. I choose a steering atom in the ligand and an “anchor” atom in a lipid in the bilayer. And the result was an impressively significant bending of the bilayer. It was pointed out in the passage above to use “the center of mass of the *entire* membrane, and the center of mass of the *entire* protein”--_000_toqmu2eabmo692bu7w6scrvb1479238207429emailandroidcom_--

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:22:36 CST