Re: Error in PMF Calculations

From: Souvik Sinha (souvik.sinha893_at_gmail.com)
Date: Wed Apr 10 2019 - 03:20:21 CDT

Does it mean, in *.colvars.traj, your colvar values are greater than 6 (for
the window 3-6 Ang)? If so, your boundary potentials are not able to keep
your molecule so close to the graphene sheet may be due to 'sterical
restriction' as Jeff mentioned earlier. Then what do you mean by this " At
the lowest point, I can find the molecules near to the surface of the
sheet"? Did you find ' 3 < COM(molecule) - COM(sheet) < 6 ' or some atom
of your bulk molecule randomly approaching surface but not the whole one?

I am not sure but increasing boundary potentials may keep your molecule in
the nearest bin to the surface, however its significance is doubtful.

On Wed, Apr 10, 2019 at 12:30 PM HEMANTH H 18310019 <hemanth.h_at_iitgn.ac.in>
wrote:

> In the bins that correspond to the closest approach (3-6) the sampling is
> zero.
>
> On Wed 10 Apr, 2019, 12:19 PM Souvik Sinha, <souvik.sinha893_at_gmail.com>
> wrote:
>
>> In your closest bin/window to the graphene sheet, does the
>> colvar updates written in *.colvars.traj restricted within your defined
>> window upper and lower boundaries?
>>
>> Souvik
>>
>> On Wed, Apr 10, 2019 at 10:14 AM HEMANTH H 18310019 <
>> hemanth.h_at_iitgn.ac.in> wrote:
>>
>>> Sir,
>>>
>>> Yes, the system has multiple layers of graphene. I am following the
>>> tutorial that was available on your webpage for setting up the PMF
>>> calculations. I was using the same input script template, across different
>>> simulations, associated with different molecules.
>>>
>>> I have set the z-transition coordinate to be the distance between the
>>> centre of mass of the residue and that of the topmost graphene sheet.
>>>
>>> At the lowest point, I can find the molecules near to the surface of the
>>> sheet, but the same is not getting registered in the count files.
>>>
>>> Thanking you,
>>> Yours faithfully,
>>> Hemanth H
>>>
>>> On Fri, Apr 5, 2019 at 7:12 PM Jeff Comer <jeffcomer_at_gmail.com> wrote:
>>>
>>>> “The sampling works for distances greater than 7 angstroms, and hence
>>>> the definition of transition coordinates must be right.”
>>>>
>>>> I don’t agree with this statement. I don’t think your transition
>>>> coordinate is defined how you are expecting.
>>>>
>>>> Do you have multiple layers of graphene? If the transition coordinate
>>>> is the center of mass of multiple layers, then you will not be able to
>>>> approach closely.
>>>>
>>>> What does system look like when the transition coordinate is at its
>>>> smallest value?
>>>>
>>>> Jeff
>>>>
>>>> On Thu, Apr 4, 2019 at 23:39 HEMANTH H 18310019 <hemanth.h_at_iitgn.ac.in>
>>>> wrote:
>>>>
>>>>> Sir,
>>>>>
>>>>> I'm using ABF to calculate the PMF for the adsorption of a Thymine
>>>>> residue. The sampling works for distances greater than 7 angstroms, and
>>>>> hence the definition of transition coordinates must be right. The sampling
>>>>> is zero in the region between 3-6 angstroms.
>>>>>
>>>>> On Thu, Apr 4, 2019 at 5:46 PM Jeff Comer <jeffcomer_at_gmail.com> wrote:
>>>>>
>>>>>> Are you using ABF? At what distance from the sheet is the sampling
>>>>>> zero? How big is your molecule? What does the PMF look like?
>>>>>>
>>>>>> My guess is that you have big bulky molecule and it is sterically
>>>>>> restricted from approaching the sheet as closely as you think it should. If
>>>>>> so, this region of the transition coordinate is very high energy and not
>>>>>> relevant.
>>>>>>
>>>>>> Another possibility is that the definition of the transition
>>>>>> coordinate (in terms of atom serial numbers), is not what you think it is.
>>>>>>
>>>>>> Jeff
>>>>>>
>>>>>> On Thu, Apr 4, 2019 at 00:23 HEMANTH H 18310019 <
>>>>>> hemanth.h_at_iitgn.ac.in> wrote:
>>>>>>
>>>>>>> Dear all,
>>>>>>>
>>>>>>> I'm trying to calculate the PMF for the absorption of a molecule on
>>>>>>> a graphene sheet. The calculation is run for 100 ns, but the sampling in
>>>>>>> the region near the sheet is zero or very negligible even after running the
>>>>>>> simulation for an extended time ( ~200 ns). I also tried setting the window
>>>>>>> size to the region that was not being sampled, but that also didn't help.
>>>>>>> Any help in this regard would be highly appreciated.
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> *Hemanth H 18310019*
>>>>>>> Research Scholar
>>>>>>> IIT Gandhinagar
>>>>>>>
>>>>>> --
>>>>>>
>>>>>> –––––––––––––––––––––––––––––––––––———————
>>>>>> Jeffrey Comer, PhD
>>>>>> Assistant Professor
>>>>>> Institute of Computational Comparative Medicine
>>>>>> Nanotechnology Innovation Center of Kansas State
>>>>>> Kansas State University
>>>>>> Office: P-213 Mosier Hall
>>>>>> Phone: 785-532-6311
>>>>>> Website: http://jeffcomer.us
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> *Hemanth H 18310019*
>>>>> Research Scholar
>>>>> IIT Gandhinagar
>>>>>
>>>> --
>>>>
>>>> –––––––––––––––––––––––––––––––––––———————
>>>> Jeffrey Comer, PhD
>>>> Assistant Professor
>>>> Institute of Computational Comparative Medicine
>>>> Nanotechnology Innovation Center of Kansas State
>>>> Kansas State University
>>>> Office: P-213 Mosier Hall
>>>> Phone: 785-532-6311
>>>> Website: http://jeffcomer.us
>>>>
>>>
>>>
>>> --
>>> *Hemanth H 18310019*
>>> Research Scholar
>>> IIT Gandhinagar
>>>
>>
>>
>> --
>> Souvik Sinha
>> Research Fellow
>> Bioinformatics Centre (SGD LAB)
>> Bose Institute
>>
>> Contact: 033 25693275
>>
>

-- 
Souvik Sinha
Research Fellow
Bioinformatics Centre (SGD LAB)
Bose Institute
Contact: 033 25693275

This archive was generated by hypermail 2.1.6 : Fri Dec 06 2019 - 23:20:35 CST