Re: [AMBER] compatibility of lipid14 with AMBER12 and lipid11/14 with NAMD v.2.10

From: Jason Swails <>
Date: Tue, 3 Jun 2014 17:58:11 -0400

On Tue, Jun 3, 2014 at 5:30 PM, Thomas Evangelidis <>

> Greetings,
> Is lipid14 fully compatible with AMBER12 for production runs?

​Yes. The functional form is equivalent to that of the "standard" Amber
force fields. [1]

> In addition are lipid11 and lipid14 compatible with NAMD v2.10?

​It should be.

> I am doing
> a simulation using lipid11 in NAMD but I 've heard from a colleague that
> the electrostatic scaling factor (scee) and the van der Waals scaling
> factor (scnb) are not computed in the same way as in AMBER.

​"Back in the day", (Amber 10 and earlier), the electrostatic and van der
Waals 1-4 scaling factors were specified in the input file -- one value for
all 1-4 interactions. Since the GLYCAM force field uses scaling factors of
1, it made mixed protein-carbohydrate systems impossible to simulate
"correctly".​ As a result, 1-4 scaling factors specific to each dihedral
was stored in the prmtop file instead.

Other codes that did not keep up with the addition to our topology file
probably still ask for both scee and scnb in the input file. I have no
idea where NAMD stands on this (you would have to ask their mailing list).
 However, as long as you don't want to mix in carbohydrates via the GLYCAM
force field, just put scee=1.2 and scnb=2.0 and be done with it.

He didn't know
> exactly though what happens. Aren't these parameters defined explicitly in
> the .prmtop file so that NAMD could know how to compute the vdW and
> electrostatics correctly, or does the code override them?

​My guess is that NAMD ignores the SCEE_SCALE_FACTOR and SCNB_SCALE_FACTOR
sections of the prmtop and simply uses scee and scnb as input by the user.
 But that may have changed in recent versions.

> And one final thing, rather irrelevant to the above: are there any
> pre-equilibrated lipid14 or lipid11 mebranes with cholesterole?

​No idea.


[1] With very few exceptions (I can think of 2), new force fields are
compatible with older versions of Amber at least as far back as Amber 9,
and most likely even further back. The 2 exceptions are ff14ipq which is
not compatible with pmemd.cuda from Amber 12 or earlier (but is perfectly
compatible with pmemd(.MPI) and sander(.MPI)) and the mixed GLYCAM-protein
simulations in Amber 10 and earlier due to the 1-4 scaling factors I
described above.

Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Tue Jun 03 2014 - 15:30:02 PDT
Custom Search