[gmx-users] Doubt about gmx wham analysis

[Justin Lemkul]

On 5/31/17 10:49 AM, Varvdekar Bhagyesh Rajendra wrote:
> Dear Justin,
> 
> During the pulling part of the umbrella sampling for finding binding affinity of the Protien-ligand system, the ligand(peptide) is deformed and its helices straighten along with large conformational changes in Protein. This is when the pull_coord1_dim = Y Y Y. But when I had used pull_coord1_dim = N N Y for one of the systems the peptide helices didn't deform. Is it reasonable for the ligand helices to deform when finding properties like Binding affinity?
> 
> Can it be avoided if the ligand orientation is not known and using pull_coord1_dim = Y Y Y is necessary.
> 

Very simply, you should be using pull_coord1_dim = Y Y Y.  That is correct.  The 
fact that you don't see a deformation with N N Y is irrelevant and perhaps 
coincidental.

Short peptides are largely unstructured in solution, and many only acquire a 
formed secondary structure upon binding to a larger protein.  This is a well 
known phenomenon.  Spurious instability of helices is a known issue with some 
parameter sets (e.g. GROMOS96 53A6 under-stabilizes helices) but with other 
force fields partial or full unfolding is not indicative of a problem.  You just 
need to be sure you're adequately sampling this conformational ensemble, which 
can be expensive.

-Justin

> Best Regards,
> Bhagyesh
> 
> ----- Original Message -----
> From: "Varvdekar Bhagyesh Rajendra" <bhagyesh.varvdekar at research.iiit.ac.in>
> To: gmx-users at gromacs.org
> Sent: Wednesday, May 31, 2017 6:42:41 PM
> Subject: Re: [gmx-users] Doubt about gmx wham analysis
> 
> Dear Justin,
> 
> Many Thanks for the swift reply and the help in answering my doubts.
> 
> Best Regards,
> Bhagyesh
> 
> ----- Original Message -----
> From: "Justin Lemkul" <jalemkul at vt.edu>
> To: gmx-users at gromacs.org
> Sent: Wednesday, May 31, 2017 6:06:59 PM
> Subject: Re: [gmx-users] Doubt about gmx wham analysis
> 
> On 5/31/17 8:34 AM, Varvdekar Bhagyesh Rajendra wrote:
>> Dear Justin,
>>
>> In my Protein-ligand system, I have used pull_coord1_dim = Y Y Y and the ligand is pulled along the COM of two groups protein-ligand in all directions to calculate binding affinity using umbrella sampling. On the other hand, the tutorials use pull_coord1_dim = N N Y. Hence, I concluded my reaction coordinate differs as mentioned in the help menu of gmx wham : "If you have some unusual  reaction coordinate you may also generate your own .pdo files and feed them with the -ip option into to gmx wham"
>>
> 
> The fact that your reaction coordinate is different from the tutorial is
> irrelevant.  You have an absolutely normal reaction coordinate and does not
> apply to what the gmx wham help text is talking about (which is probably a very
> outdated note at this point, given the massive changes in the pull code in
> recent versions).
> 
>> Also, the following warning is thrown by gmx wham: " WARNING, no data point in bin 7 (z=0.403502) ! You may not get a reasonable profile. Check your histograms! "
>> I was not sure if the z corresponds to the coordinate axis or just the z-axis (which was not the only reaction coordinate in my system).
>>
> 
> It is the length along zeta, the reaction coordinate.  Don't confuse it with the
> z-axis, it's just shorthand.
> 
>> All this made me conclude that the pullx files are not enough and the so called pdo files must be necessary, hence the doubt arised. I would appreciate if some more light is shed in this area.
>>
> 
> You need better sampling, not archaic file formats.
> 
> -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

==================================================