Re: Ficoll/Dextran CHARMM Force Field

From: Michelle Kuttel (
Date: Wed Aug 24 2022 - 02:48:36 CDT

Hi Angelo

It’s not advertised, but you can use CarbBuilder for this, available from*mkuttel/Downloads.html__;fg!!DZ3fjg!8NsRk3E2L11FXSUCMjgIckciws5bmUZd5n13Cp5X4Jx0awDaco2fufympNzJVkYMSpWNkUIZ-W63rQqY475CN_A$ <*mkuttel/Downloads.html__;fg!!DZ3fjg!8NsRk3E2L11FXSUCMjgIckciws5bmUZd5n13Cp5X4Jx0awDaco2fufympNzJVkYMSpWNkUIZ-W63rQqY475CN_A$ >
Use the “-PSF” flag to generate the CHARMM .psf files. You will need the namd psfgen.exe program in the root CarbBuilder directory (the one released with CarbBuilder is the 32 bit Mac version, which will only work if you are on Mac. It has to be named psfgen.exe - even on Linux/MAc… sorry!).

You can create different dextran chains, e.g.
mono ../CarbBuilder2.exe -i "->6)aDGlc(1->3)aDGlc(1->" -r 100 -PSF -o dextran6
This will give you a chain of 100 repeats of the ->6)aDGlc(1->3)aDGlc(1-> repeat unit (i.e. 200 residue chain).

The chains will then need to be combined into a simulation system with water - VMD can do this for you, with a bit of effort.
I am happy to give guidance on how to do this - please email me directly.


Michelle Kuttel
Associate Professor
Computer Science Department
University of Cape Town
+27 21 6505107

> On 23 Aug 2022, at 19:58, Angelo Rossi <> wrote:
> Jeff Comer is correct that the CHARMM carbohydrate force field is much more accurate for Dextran 70.
> What I thought is that one could begin by running a series of small simulations with different size dextran oligomers replicated to provide the idea of crowding. One could ramp up the number of dextran oligomers to observe property differences. The longer the polymer, correspondingly longer equilibration times are needed. The 389 glucose monomers per polymer times the total number of polymer chains could begin to be quite significant for gleaning property evaluation. The number of Dextran 70 polymer chains required for crowding isn't mentioned.
> Kind regards,
> Angelo
> On Tue, Aug 23, 2022 at 1:13 PM Jeff Comer < <>> wrote:
> Dextran is not a well defined species, but a polymer of glucose with
> mainly alpha 1,3 and alpha 1,6 linkages (
>;!!DZ3fjg!4oN-m8V3HtruppgcFZ88vebTiMLK6mk3PqoPXYmlNn_fOsjqEyGcSJG-L1VAH7JvVFYvXRZROy35RL3gGb4$ <;!!DZ3fjg!4oN-m8V3HtruppgcFZ88vebTiMLK6mk3PqoPXYmlNn_fOsjqEyGcSJG-L1VAH7JvVFYvXRZROy35RL3gGb4$> ). If you don't know what that
> means, I recommend studying a bit about carbohydrate structure.
> Dextran 70 refers to Dextran polymers with molecular masses of about
> 70 kDa. So we are talking about 70000/180.156 ~ 389 glucose units.
> CGenFF, as suggested by Angelo Rossi, is probably the easiest
> solution, but not a very correct one for carbohydrates like dextran
> and ficoll. The CHARMM carbohydrate force field would likely do a
> better job than CGenFF and be the preferred solution.
> To make an accurate model of dextran 70, you'll need to connect ~300
> glucose units. Alpha glucose is resname AGLC in the official CHARMM
> carbohydrate force field topology file top_all36_carb.rtf. You then
> will patch them together using the α1->6 linkage and α1->3 linkage
> patches in the same file (16AT and 13aa, I think).
> There are several different patches that are actually the same in
> topology (13aa, 13ab, 13ba, 13bb) or (16at, 16bt). The only difference
> is the automatically generated coordinates. It does not matter which
> you use if you set the coordinates correctly yourself.
> Regards,
> Jeff
> –––––––––––––––––––––––––––––––––––———————
> Jeffrey Comer, PhD
> Associate Professor
> Department of Anatomy and Physiology
> Kansas State University
> Office: 334 Coles Hall
> Phone: 785-532-6311
> Website:;!!DZ3fjg!4oN-m8V3HtruppgcFZ88vebTiMLK6mk3PqoPXYmlNn_fOsjqEyGcSJG-L1VAH7JvVFYvXRZROy35VKmES0Y$ <;!!DZ3fjg!4oN-m8V3HtruppgcFZ88vebTiMLK6mk3PqoPXYmlNn_fOsjqEyGcSJG-L1VAH7JvVFYvXRZROy35VKmES0Y$>
> On Mon, Aug 22, 2022 at 2:19 PM Polly Ren < <>> wrote:
> >
> > Hello!
> >
> > I was wondering whether anyone has the CHARMM force field for the Ficoll 70 or Dextran 70 molecules? We are trying to perform a macromolecule crowding simulation but we are unable to find the force field for either polymer. Any guidance on where they could be found or how they can be constructed would be greatly appreciated!
> >
> > Thank you,
> > Polly Ren

This archive was generated by hypermail 2.1.6 : Tue Dec 13 2022 - 14:32:44 CST