[gmx-users] Distance restraint and pull code

Fri Jun 10 03:45:13 CEST 2016

hi,
I included this bond assignment right after [bonds] in the topology.
[ bonds ]
;   i     j  func       b0      kb
 1354  3852      6      0.21    4000
 1390  3852      6      0.21    4000
 1721  3852      6      0.21    4000
 1520  3852      6      0.40    4000

It seems the type 6 works now, 3/4 pairs are restrainted, 1 is not (the
first pair: 1354  3852      6      0.21    4000). This pair keeps
fluatuating around ~3.5 A.
To keep it, I increase the kb of first pair from 4000 to 6000, then the
result is better than the old one.
Both simulations are added harmonic potential, same b0 and kb value, but
pull code works more effectively than distance restraint.
What do you think about this?
I think the reason is the pull code just applies pull force to the an atom,
to keep the restrainted distance with the other. Differently, the force is
applied to both atoms in [bonds].
Anw, thanks for your help so much. :D
HT

On Fri, Jun 10, 2016 at 3:52 AM, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 6/9/16 11:21 AM, Hong Tham wrote:
>
>> hi Justin,
>> in the manual, the [distance restraint] will add flat-bottom harmonic
>> potential to keep distance between 2 atoms. The low, up1, up2 is
>> corresponds to ro, r1, r2 in the equation 4.83 (manual 5.0.7).
>> I understand if rij < r low,   up1 <rij > up2, mdrun will add the
>> harmonic potential, if rij > up2 , the linear potential will be added. no
>> potential is added if low < rij < r up1.
>> However, because the pull code will add harmonic potential, I did think
>> the distance restraint should do the same thing ( harmonic restraint
>> instead of flat-bottom harmonic restraint). I am compare between 2 way in
>> maintaining the zn coordination. Therefore, I did set the r low = r up1, r
>> up 2 very big. I intend to let mdrun recognize if rij < r low = r up1 or
>> rij > r up2, the harmonic potential will be added. I expect that the
>> distance rij between 2 atoms can not reach that high r up2. and I think,
>> after modifying, the 4.83 equation will look like 4.31 (harmonic potential).
>> even, if i run genrestr to generate the distance restraint itp which is
>> later included in the topology file, it return the [distance restraint]
>> with low = up1, the disre-up2 is the up2 value.
>> As your reply, that mean I can not do like that? Gromacs can not
>> recognize that, and it will do the way as you mentioned? I don’t want that,
>> actually. So, If I want to apply distance restraint, just simply use
>> flat-bottom potential?
>> Sorry if I did ask like that, I’m still studying to understand more about
>> Gromacs and how it works. and sometime, I may misunderstand. :D
>>
>
> If you want a plain harmonic potential between different atoms, use bond
> type 6 (does not generate exclusions, not subject to constraints).
>
> -Justin
>
>
> Thanks so much ^^
>> HT
>>
>>
>> On Jun 9, 2016, at 9:58 PM, Justin Lemkul <jalemkul at vt.edu> wrote:
>>>
>>>
>>>
>>> On 6/9/16 8:54 AM, HongTham wrote:
>>>
>>>> Hi,
>>>> I have another question. I applied 4 pull codes (3 to keep Zn ion in
>>>> coordination with 3 HIS, 1 to keep Zn ion in coordination with the
>>>> ligand).
>>>> The pull code worked effectively.
>>>> With the same system, when applying 3 distance restraints (to keep Zn
>>>> ion
>>>> in coordination with 3 HIS) and 1 pull code to keep the Zn ion in
>>>> coordination with the ligand. However, the distance restraints didn't
>>>> work.
>>>> The distance restraints and pull code for restrains in principle both
>>>> work
>>>> in the same manner. In both cases,  a harmonic potential is added to
>>>> maintain the distance between 2 selections.
>>>> However, there is an inconsistency between the two approaches.
>>>> This is the distance restaints code:
>>>> [ distance_restraints ]
>>>> ;   i     j       type  label      funct    lo        up1        up2
>>>> weight
>>>> 1354  3852 1     1         1   0.21       0.21    5.21             4
>>>> 1390  3852 1     2         1   0.21       0.21    5.21             4
>>>> 1721  3852 1     3         1   0.21       0.21    5.21             4
>>>> 1520  3852 1     4         1   0.40       0.40    5.40             2
>>>>
>>>> which is included in the topology file.
>>>>
>>>> The pull codes used are similar to the previous, only changed the
>>>> pull-coord-k parameter(s) to be the same with the distance restraints.
>>>> pull-coord1-k = 4000 ; kJ mol^-1 nm^-2
>>>> pull-coord2-k = 4000 ; kJ mol^-1 nm^-2
>>>> pull-coord3-k = 4000 ; kJ mol^-1 nm^-2
>>>> pull-coord4-k = 2000 ; kJ mol^-1 nm^-2
>>>>
>>>> The question is, why is there an inconsistency? The distance between the
>>>> histidines and the Zn ion keeps increasing as the simulation is
>>>> progressing, if distance restraints are applied. If only pull codes are
>>>> used, then everything is fine. All the distances converge well.
>>>>
>>>
>>> Your distance restraints make no sense.  Look at figure 4.15 in the
>>> manual. You're telling mdrun "apply zero biasing potential from 0.21 - 5.21
>>> (or 0.4 - 5.4) nm), then try to make the potential steeper from 5.21 - 4 nm
>>> (or 5.4 - 2 nm).  So mdrun does what you tell it; essentially no restraint
>>> is applied unless the distances are massive.
>>>
>>> -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.
>>>
>>
>>
> --
> ==================================================
>
> 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.
>