Re: Dielectric constant for MDFF of a membrane protein.

From: JC Gumbart (
Date: Fri Mar 30 2012 - 13:55:56 CDT

That's a really good question. Within a protein itself the dielectric is also not 80 (I think 4-5?) though, so even for soluble proteins it's a bad approximation, particularly for very short-range interactions. As for the membrane, since the protein is fairly compact, it's unlikely that there are any significant intra-protein interactions that would have a lot of membrane in the intervening space, vs. protein or water. So I suspect using a much smaller dielectric in the hopes of mimicking the membrane is actually even less justified.

For what it's worth, in the past, I've just used 80 for fitting of membrane proteins. Most everything is driven by local interactions anyway, so the point is just to prevent more distant electrostatics from dominating. But the conventual wisdom going forward is that one should carry out multiple rounds of fitting, first in vacuum and then in the proper water or water/membrane environment.

Here's an example of a paper where fitting was done with a membrane present (but not coupled to the map):


On Mar 30, 2012, at 9:18 AM, Raul Araya wrote:

> Dear NAMD users
> I´m triyinf to perform a MDFF of a membrane channel protein using an
> available Cryo-EM density map. Everything seems quite straightforward
> from the NAMD-MDFF tutorial, but there is something that is not so
> easy to define. In the typical MDFF procedure a dielectric constant
> must be set to represent the solvent, and thus for water soluble
> proteins this value is 80...But in the case of a transmembrane
> channel, should I use a membrane like dielectric (i.e 4-6)?? or should
> I start by using 80 but keeping secondary stricture highly restrained?
> Thanks.
> Raúl Araya Secchi
> B.Sc Molecular Biotechnology.
> Molecular Biotechnology Engineer.
> PhD Student (Biotechnology Program. UNAB, Chile)
> Computational Biology Lab (DLab)
> Center for Mathematical Modeling (CMM)
> Facultad de Ciencias Físicas y Matemáticas.
> Universidad de Chile.

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:21:49 CST