Re: [AMBER] 3D-RISM solvation free energy of a molecule in an organic solvent (not water)

From: Smith, Micholas D. <smithmd.ornl.gov>
Date: Mon, 19 Mar 2018 17:59:09 +0000

Thanks Tyler. That cleared up a lot of the issues.



===================
Micholas Dean Smith, PhD. MRSC
Post-doctoral Research Associate
University of Tennessee/Oak Ridge National Laboratory
Center for Molecular Biophysics

________________________________________
From: Tyler Luchko (Lists) <tluchko.lists.gmail.com>
Sent: Monday, March 19, 2018 1:53 PM
To: AMBER Mailing List
Subject: Re: [AMBER] 3D-RISM solvation free energy of a molecule in an organic solvent (not water)

Hi Micholas,

To use a different solvent with 3D-RISM, you need to generate a new XVV file to pass to rism3d.snglpnt. Please see the required --xvv flag.

An XVV file is typically created with rism1d, which requires an MDL file that contains the model of your solvent molecule. The format of the MDL file is in the manual (p. 118), as are instructions for using rism1d.

rism1d also requires a .inp file to specify details of the numeric solver and the solvent. It is here that you specify the temperature and density of the solvent. It is possible to have more than one type of solvent, each with its own MDL file. In this case you must specify the density of each solvent.

If you are planning on using correction methods like the Universal Correction (UC) or 3D-RISM-PC+, please keep in mind that these haven’t been tested for large solutes yet. Although UC does work well for non-polar solvents, it has to be parameterized for the specific solvent and XVV file you are interested in using. Obtaining data for mixtures of solvents to fit against may be difficult. PC+ is known to work well for water but there have been mixed results for for other liquids. I found it worked poorly for cyclohexane

Luchko, Blinov, Limon, Joyce, & Kovalenko. (2016). J Comput Aided Mol Des, 1–13.

but other have found it to give reasonable results for course grained models of a number of solvents

Misin, Palmer, & Fedorov. (2016). J. Phys. Chem. B, 120, 5724–5731.

Also, PC+ requires PSE-3 or HNC closures, which may be difficult to converge. In this case, a closure bootstrap may help (see 7.3.1.1 of the manual).

Hope this helps,

Tyler

> On Mar 19, 2018, at 7:40 AM, Smith, Micholas D. <smithmd.ornl.gov> wrote:
>
> Greetings,
>
> I am trying to make use of the 3D-RISM module to calculate the solvation free-energy of protein in an organic solvent, rather than water. I have been going over the manual, but can't quite make sense of what I need to modify in the input to make this work since it seems rismsnlpoint.py is hard-coded to use either TIP3P or spc/e. Any ideas on how to proceed?
>
> As a follow up question: what about the case of a cosolvent mixture? Say a mixture of ethanol and water?
>
> Thanks for your help.
>
> -Micholas
>
> ===================
> Micholas Dean Smith, PhD. MRSC
> Post-doctoral Research Associate
> University of Tennessee/Oak Ridge National Laboratory
> Center for Molecular Biophysics
>
>
>
> ===================
> Micholas Dean Smith, PhD. MRSC
> Post-doctoral Research Associate
> University of Tennessee/Oak Ridge National Laboratory
> Center for Molecular Biophysics
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber

_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Mar 19 2018 - 11:00:04 PDT
Custom Search