Re: Defining custom Colvars to measure pmf with Umbrella Sampling

From: Hemanth Vemuri (hemanthvemuri1997_at_gmail.com)
Date: Thu Jun 22 2017 - 00:26:31 CDT

My initial approach to the problem was along similar lines as suggested by
you, i thought of visualizing the simulations in vmd to check where the
penetrated waters were clustering and then placing dummy atoms at those
locations and taking the water coordination number of this group of dummy
atoms as the reaction coordinated for Umbrella sampling. However , then i
could not figure out in which locations should the dummy atoms be placed in
the protein. Then i thought of doing exactly the same thing as suggested by
you , that is using the coordNum between protein and water but then again
as you mentioned it will include "reasonably consistent "shell" term of
protein water contacts" , i had earlier defined these waters as surface
waters and are defined as the contact water which are not penetrating
water. Is there any way in which i can not include these surface water in
my Umbrella sampling run?. Since i wanted to find a way of not including
these surface waters in my goal of finding the variation in pmf as number
of penetrated water , i thought maybe defining a custom colvars might help.

Thanks for advice

Hemanth Vemuri,
undergraduate student
Center for Computational Natural Sciences and Bioinformatics,
International Institute of Information Technology,
Gachibowli, Hyderabad 500 032
Telangana, India

On Wed, Jun 21, 2017 at 9:36 PM, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov>
wrote:

> As an approximation, something like the contacts (coordnum) between
> protein heavy atoms and water heavy atoms will give the same sort of flavor
> as the number of intercalated waters. There will be a reasonably consistent
> "shell" term of protein water contacts you can't do anything about, but
> depending on the protein configuration, the number of waters within your
> protein would change by a measurable amount.
>
> -Josh
>
> On 06/20/2017 08:56 PM, Giacomo Fiorin wrote:
> Hello Hemanth, this is not a trivial problem, because your definition
> makes sense for analysis (it is pretty much a VMD selection) but in
> dynamics you need to use continuous forces.
>
> There are a few approaches to this (see e.g. one below), but first you
> have to ask yourself what is the role of protein flexibility in your
> problem.
>
> Giacomo
>
> http://www.pnas.org/content/112/27/8181.
> safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.
> pnas.org%2Fcontent%2F112%2F27%2F8181&data=02%7C01%7CJoshua.
> Vermaas%40nrel.gov%7C67f7352495e0416d8f4d08d4b850e3c5%
> 7Ca0f29d7e28cd4f5484427885aee7c080%7C1%7C0%7C636336104941366296&sdata=v%
> 2FQdIr9pX7veLbt9RwYasKMux%2BP%2B9vqoqqnURZx6MDA%3D&reserved=0>
>
> On Tue, Jun 20, 2017 at 1:25 AM, Hemanth Vemuri <
>
hemanthvemuri1997_at_gmail.com<mailto:hemanthvemuri1997_at_gmail.com>> wrote:
> Hello,
> My goal is to measure the changes in pmf as we vary the number of water
> molecules inside/around(1) a protein. But the problem i am facing is that
> there is no suitable colvars which will help me achieve this, like for
> example if i want to measure pmf with radius of gyration as the reaction
> coordinate , there is already Rg collective variable which i can specify
> in the colvars input file, but there is no colvar with which i can measure
> the pmf with number of water as a reaction coordindate. Hence i wanted to
> know if there is a way with which i can define a custom collective variable
> or is there any other way to reach my goal.
>
> (1) - I call these waters as penetrated waters and am defining them as
> waters belonging to contact water which are not within 3 angstrom of bulk
> water.
> Contact water are defined as water within 3.5 angstrom of protein and the
> water which are not contact water are defined as bulk water.
>
> Thanks for advice
>
> Hemanth Vemuri,
> undergraduate student
> Center for Computational Natural Sciences and Bioinformatics,
> International Institute of Information Technology,
> Gachibowli, Hyderabad 500 032
> Telangana, India
>
>
>
>
> --
> Giacomo Fiorin
> Associate Professor of Research, Temple University, Philadelphia, PA
> Contractor, National Institutes of Health, Bethesda, MD
> http://goo.gl/Q3TBQU.
> outlook.com/?url=http%3A%2F%2Fgoo.gl%2FQ3TBQU&data=02%
> 7C01%7CJoshua.Vermaas%40nrel.gov%7C67f7352495e0416d8f4d08d4b850e3c5%
> 7Ca0f29d7e28cd4f5484427885aee7c080%7C1%7C0%7C636336104941366296&sdata=
> 8TUc8Q%2B8lBY8CUA3yejF3kSuciU3WVXjzZ%2FaWeMD%2BLE%3D&reserved=0>
>
https://github.com/giacomofiorin.
> safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.
> com%2Fgiacomofiorin&data=02%7C01%7CJoshua.Vermaas%40nrel.gov%
> 7C67f7352495e0416d8f4d08d4b850e3c5%7Ca0f29d7e28cd4f5484427885aee7
> c080%7C1%7C0%7C636336104941366296&sdata=5%2FxkPzxcq%
> 2FPLaxL5RyHSqTdXdAJCrbQUCngVDSwKT2o%3D&reserved=0>
>
>

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:20:23 CST