[gmx-users] Doubt about Free Energy Calculations using g_bar

Mon Jun 12 02:11:19 CEST 2017

Dear Bhagyesh,

It is really nice that you have a reference to a published paper that 
features language like "enthalpically stabilized" in its abstract. ;) I 
also think that fundamentally it is indeed reasonable to believe that 
phonons (a sane term for normal modes) and their spectra for different 
protein conformations should be different.

However, Mark's point is along the lines of observing those differences 
in an MD simulation and then correctly attributing them to what changes 
have taken place upon ligand binding, let alone extending this to the 
actual physics that takes place in an experiment. It is a bit 
unfortunate that the biophysics community borrowed studying phonons from 
solid-state physics without a deep understanding of the key differences 
between solid crystals and floppy overdamped molecules surrounded by 
water, especially when it comes to low-frequency modes. Gromacs has some 
tools to calculate the corresponding Hessian matrix elements, etc (e.g. 
see http://lorentz.immstr.pasteur.fr/gromacs/). In addition, you could 
script your own trajectory analyzer to plot the frequency of states, as 
well as calculate things like the Debye-Waller factor for selected atoms 
in particular regions of your protein. I would even bet that you could 
see differences in your plots for various protein configurations. 
However, here, unlike in crystals, the system is fundamentally an 
overdamped mess, more so when using bond constraints. Things produced by 
SHAKE and LINCS do not really lend themselves very to phonons.

This essentially suggests exercising great care when interpreting the 
data. Not sure how helpful this is, but this is just a general comment. 
As far as being able to accomplish such an analysis is Gromacs, you most 
certainly are. The only question is whether it is useful to anyone, 
aside from publishing a paper.

Alex

On 6/11/2017 3:48 PM, Varvdekar Bhagyesh Rajendra wrote:
> Dear Mark,
>
> I humbly disagree that the observations could not provide an insight into the physical systems because a previous the work provides a reasonable proof: "Theory and normal-mode analysis of change in protein vibrational dynamics on ligand binding", DOI: 10.1021/jp909677p
>
> I wish to extend the said work and it speaks about the generation of a hypothetical “intermediate” state, without the ligand force field but with a structure resembling to that of the Bound complex state. It reveals that during the binding occurs from intermediate to the bound state, the vibrational entropies of both the protein and the ligand decrease, thereby predicting the physical state properties.
>
> Even if the hypothesis is not true, I wish to proceed the required procedure and it would be helpful if you could provide any valuable revelation of doing the same in gromacs.
>
> Best Regards,
>
> Bhagyesh
>
> ----- Original Message -----
> From: "Mark Abraham" <mark.j.abraham at gmail.com>
> To: gmx-users at gromacs.org
> Sent: Monday, June 12, 2017 2:43:53 AM
> Subject: Re: [gmx-users] Doubt about Free Energy Calculations using g_bar
>
> Hi,
>
> On Sun, Jun 11, 2017 at 10:39 PM Varvdekar Bhagyesh Rajendra <
> bhagyesh.varvdekar at research.iiit.ac.in> wrote:
>
>> Dear Mark,
>>
>> Unless I'm mistaken, even though they are non-physical states and do not
>> exist in reality, they have the required modification in the actual
>> physical state (precisely, the non bonded interaction between protein and
>> ligand) that I actually need and I wish to use the same and do Normal mode
>> calculations to study the effect of the said perturbation at each lambda
>> from 0 to 1.
>>
>> I would appreciate any suggestions to undertake the above quest.
>>
> I expect that the numbers you seek will be computed. But you're also going
> to scale down e.g. the ligand-solute interactions, and IMO it's unlikely
> you'll observe anything that you can claim is insight into the physical
> systems.
>
> Mark
>
>
>> Best Regards,
>>
>> Bhagyesh
>>
>> ----- Original Message -----
>> From: "Mark Abraham" <mark.j.abraham at gmail.com>
>> To: gmx-users at gromacs.org
>> Sent: Sunday, June 11, 2017 9:22:51 PM
>> Subject: Re: [gmx-users] Doubt about Free Energy Calculations using g_bar
>>
>> Hi,
>>
>> To what end? These are non-physical states.
>>
>> Mark
>>
>> On Sun, 11 Jun 2017 16:28 Varvdekar Bhagyesh Rajendra <
>> bhagyesh.varvdekar at research.iiit.ac.in> wrote:
>>
>>> Dear Justin,
>>>
>>> Thank you for the valuable insight.
>>>
>>> While on the same subject, can you give any comments on the possibility
>> of
>>> using the same Free energy code in Gromacs to scale the non bonded
>>> interactions between protein & ligand while simultaneously doing Normal
>>> Mode analysis on the scaled interaction conformation to study the effect
>> of
>>> the scaled interactions.
>>>
>>> If its not plausible in Gromacs, can you suggest any other software.
>>>
>>> Best Regards,
>>>
>>> Bhagyesh
>>>
>>> ----- Original Message -----
>>> From: "Justin Lemkul" <jalemkul at vt.edu>
>>> To: gmx-users at gromacs.org
>>> Sent: Sunday, June 11, 2017 7:01:26 AM
>>> Subject: Re: [gmx-users] Doubt about Free Energy Calculations using g_bar
>>>
>>> On 6/10/17 6:15 AM, Varvdekar Bhagyesh Rajendra wrote:
>>>> Dear Justin,
>>>>
>>>> Just so we are on the same page, this means that if I don't touch the
>>> topology file and use the following mdp snippet, charges *are present* in
>>> the topology file for the ligand group and they are automatically set to
>>> zero while turning on vdW interactions (from 0 to 1) right ? Hence, there
>>> is no need to manually set charges to zero (The old style of doing the
>>> calculations), right?.
>>> Yes.  You can easily confirm this using energygrps between your
>> transformed
>>> molecule and the rest of the system.  Coul-SR should be zero.
>>>
>>>> Then the charges too, are turned off from 0 to 1? (1 state being the
>>> actual charges present in the topology file).
>>> Charges are turned *on* if they are defined as off in the A-state
>>> (lambda=0) and
>>> on in the B-state (lambda=1).
>>>
>>> -Justin
>>>
>>> --
>>> ==================================================
>>>
>>> Justin A. Lemkul, Ph.D.
>>> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>>>
>>> Department of Pharmaceutical Sciences
>>> School of Pharmacy
>>> Health Sciences Facility II, Room 629
>>> University of Maryland, Baltimore
>>> 20 Penn St.
>>> Baltimore, MD 21201
>>>
>>> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
>>> http://mackerell.umaryland.edu/~jalemkul
>>>
>>> ==================================================
>>> --
>>> Gromacs Users mailing list
>>>
>>> * Please search the archive at
>>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>>> posting!
>>>
>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>
>>> * For (un)subscribe requests visit
>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>>> send a mail to gmx-users-request at gromacs.org.
>>> --
>>> Gromacs Users mailing list
>>>
>>> * Please search the archive at
>>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>>> posting!
>>>
>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>
>>> * For (un)subscribe requests visit
>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>>> send a mail to gmx-users-request at gromacs.org.
>>>
>> --
>> Gromacs Users mailing list
>>
>> * Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>> posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>> send a mail to gmx-users-request at gromacs.org.
>> --
>> Gromacs Users mailing list
>>
>> * Please search the archive at
>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
>> posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>> send a mail to gmx-users-request at gromacs.org.
>>