Re: Centre of mass distance in ABF

From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Thu Dec 17 2020 - 09:03:45 CST

Once again, please keep the list in CC. Sharing this information is the
primary reason why the list exists in the first place.

It is hard to make a good guess about the process without having access to
your data (i.e. without being you).

You have defined wall restraints at the boundaries of each window. First,
be aware of the following:
https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:colvars_config_changes
and perhaps consider making them explicit harmonicWalls restraints.

Second, those restraints ought to work if the distance is computed
correctly. Does the value of that distance make sense to you when you look
at the trajectory? Look into the role of PBCs:
https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:colvar_atom_groups_wrapping

On Thu, Dec 17, 2020 at 9:53 AM Aashish Bhatt <aashish.ph16221_at_inst.ac.in>
wrote:

> Dear sir,
>
> I want to perform the simulation of the formation of inclusion complex in
> six windows i.e, I want to move the peptide towards the cyclodextrin cavity
> by 5Å in each window(Total 30 Å).
>
> But in my case, the whole process is getting completed within the first
> window.
>
> Herein I am pasting my colvar input file.
>
> Best regards
>
>
> colvarsTrajFrequency 100
> colvarsRestartFrequency 1000
>
> colvar {
> name COMDistance
>
> width 0.1
>
> lowerboundary -20.0
> upperboundary -15.0
>
> lowerwallconstant 10.0
> upperwallconstant 10.0
>
> distance {
> group1 {
> atomnumbers { 148 149 150 151 152 155 157 159 160 161 162 163 165
> 167 169 171 } ## Peptide atoms
> }
> group2 {
> atomnumbers { 21 42 63 84 105 126 147 } ## Cyclodextrin
> }
> }
> }
>
> abf {
> colvars COMDistance
> fullSamples 500
>
> On Thu, Dec 17, 2020 at 8:08 PM Giacomo Fiorin <giacomo.fiorin_at_gmail.com>
> wrote:
>
>> Hello Aashish, please copy the list in your reply.
>>
>> Can you please clarify what is the problem?
>>
>> On Thu, Dec 17, 2020 at 9:26 AM Aashish Bhatt <aashish.ph16221_at_inst.ac.in>
>> wrote:
>>
>>> Dear Sir
>>>
>>> Your assumption is correct.
>>> My group1 atoms are 20Å far from the group2 atoms.
>>> I have performed 10000 steps minimisation and the equilibration with
>>> solvent for 500ps.
>>> In the solvent equilibration, I have constrained the solute then
>>> performed ABF dynamics (NVT).
>>> During ABF simulation Group1 atoms are going inside the cyclodextrin
>>> molecule after the very first frame. Although, I want to perform a
>>> simulation for the formation of the inclusion complex at 6 windows
>>> instead of at the first window.
>>> Can you please tell me if there is any way to modulate the process?
>>> I have attached the colvar file with this mail.
>>> I have applied the 5kcal/mol/Å2 restrained force on the cyclodextrin
>>> (glycosidic oxygen).
>>>
>>>
>>> Best regards
>>>
>>> Aashish
>>>
>>>
>>> On Sat, Dec 12, 2020 at 8:48 PM Giacomo Fiorin <giacomo.fiorin_at_gmail.com>
>>> wrote:
>>>
>>>> Hi Asshish, you didn't say which type of variable (function) you used.
>>>> Assuming that you used a "distance" function:
>>>>
>>>> https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:cvc_distance
>>>> you are defining indeed the distance between the two COMs.
>>>>
>>>> Specifically, the "program" that computes the COMs of groups is
>>>> actually NAMD, while Colvars uses them to define the collective variable
>>>> and the biasing forces on it that allow you enhanced sampling and PMF
>>>> computation. (Instead, other types of collective variables that are not
>>>> based on COMs are computed entirely within Colvars.)
>>>>
>>>> In your publication, you should therefore cite both the NAMD and
>>>> Colvars papers:
>>>> https://doi.org/10.1063/5.0014475
>>>> https://doi.org/10.1080/00268976.2013.813594
>>>>
>>>> Thanks,
>>>> Giacomo
>>>>
>>>> On Sat, Dec 12, 2020 at 3:18 AM Aashish Bhatt <
>>>> aashish.ph16221_at_inst.ac.in> wrote:
>>>>
>>>>> Dear Sir
>>>>>
>>>>> I have a system with cyclodextrin and peptide. I want to study the
>>>>> formation of the inclusion complex via the adaptive biasing force method.
>>>>> As discussed in the following paper.
>>>>> DOI: 10.1039/c7cp02292a
>>>>> I took the group 1 colvar glycosidic bond oxygen atoms and in another
>>>>> group 2 selection of heavy atoms of the N terminal or C-terminal amino acid
>>>>> residue.
>>>>> I want to know whether the program automatically calculates the COM or
>>>>> I have to give a different format.
>>>>>
>>>>>
>>>>> Best regards
>>>>>
>>>>> Aashish
>>>>>
>>>>>
>>>>>
>>>>>

This archive was generated by hypermail 2.1.6 : Fri Dec 31 2021 - 23:17:10 CST