[gmx-users] Parametrisation of the cyclic nucleotides in Gromos force fields

James Starlight jmsstarlight at gmail.com
Sat Dec 8 12:20:54 CET 2012


Today I've performed test simulation of the protein-cAMP complex with
the charmm27 force field and obtain good stable system. I've compared
polar contacts of my complex ( ligand have been done with SwissParam)
with the X-ray structure of that protein solved with the same complex
and obtained similar picture in both cases.
When I've checked ligand's topology produced by Swiss param I noticed
that it was built from some building blocks generated by SwissParam (
and found in the same topology file) with the specified charge
distribution ( the below example is the cAMP mollecule)

; nr type resnr resid atom cgnr charge mass
   1 NPYL 1  LIG N9      1  0.0476  14.0067
   2 C5A  1  LIG C8      2  0.0365  12.0110
   3 N5B  1  LIG N7      3 -0.5653  14.0067
   4 C5B  1  LIG C5      4  0.2272  12.0110
   5 CB   1  LIG C6      5  0.4100  12.0110
   6 NPYD 1  LIG N1      6 -0.6200  14.0067
   7 CB   1  LIG C2      7  0.4700  12.0110
   8 NPYD 1  LIG N3      8 -0.5670  14.0067
   9 C5A  1  LIG C4      9  0.1054  12.0110
  10 HCMM 1  LIG H01    10  0.1500   1.0079
  11 HCMM 1  LIG H02    11  0.0000   1.0079
  12 HCMM 1  LIG H03    12  0.0000   1.0079
  13 HCMM 1  LIG H04    13  0.0000   1.0079
  14 HCMM 1  LIG H05    14  0.0000   1.0079
  15 HOR  1  LIG H06    15  0.4000   1.0079
  16 HCMM 1  LIG H07    16  0.0000   1.0079
  17 HCMM 1  LIG H08    17  0.1500   1.0079
  18 HNCO 1  LIG H09    18  0.4000   1.0079
  19 HNCO 1  LIG H10    19  0.4000   1.0079
  20 HCMM 1  LIG H11    20  0.0000   1.0079
  21 PO4  1  LIG P      21  1.4424  30.9738
  22 O2CM 1  LIG O1P    22 -0.9500  15.9994
  23 O2CM 1  LIG O2P    23 -0.9500  15.9994
  24 OR   1  LIG O5'    24 -0.5512  15.9994
  25 CR   1  LIG C5'    25  0.2800  12.0110
  26 CR   1  LIG C4'    26  0.2800  12.0110
  27 OR   1  LIG O4'    27 -0.5600  15.9994
  28 CR   1  LIG C3'    28  0.2800  12.0110
  29 OR   1  LIG O3'    29 -0.5512  15.9994
  30 CR   1  LIG C2'    30  0.2800  12.0110
  31 OR   1  LIG O2'    31 -0.6800  15.9994
  32 CR   1  LIG C1'    32  0.5356  12.0110
  33 NC=C 1  LIG N6     33 -0.9000  14.0067

where building blocks were

[ atomtypes ]
; name at.num  mass   charge  ptype    sigma            epsilon
NPYL    7   14.0067  0.0  A         0.306469    0.376560
C5A     6   12.0110  0.0  A         0.363487    0.209200
N5B     7   14.0067  0.0  A         0.329632    0.836800
C5B     6   12.0110  0.0  A         0.363487    0.209200
CB      6   12.0110  0.0  A         0.355005    0.292880
NPYD    7   14.0067  0.0  A         0.329632    0.836800
HCMM    1    1.0079  0.0  A         0.235197    0.092048
HOR     1    1.0079  0.0  A         0.040001    0.192464
HNCO    1    1.0079  0.0  A         0.040001    0.192464
PO4    15   30.9738  0.0  A         0.383086    2.447640
O2CM    8   15.9994  0.0  A         0.302905    0.502080
OR      8   15.9994  0.0  A         0.315378    0.636386
CR      6   12.0110  0.0  A         0.387541    0.230120
NC=C    7   14.0067  0.0  A         0.329632    0.836800

So its interesting for me
1- on what assumptions that blocks were generated ?
2- Why charge in [ atomtypes ]  (zero) differes from  the charges in
the topology of the same groups? how I can check correctness of charge
distribution in such itp files ?
3- What is sigma     and        epsilon in the [ atomtypes ] ? :)
4- Have anybody else used Swiss param for modeling protein-ligands
systems? Might it be used with the charmm36 set ?

James

2012/12/7, Justin Lemkul <jalemkul at vt.edu>:
>
>
> On 12/7/12 2:21 PM, James Starlight wrote:
>> Justin,
>>
>> with that charmm27 cutoffs (rlist=1.2 rlistlong=1.4 rcoulomb=1.2 rvdw=1.2
>> rvdw_switch=0.8 and vdwtype=switch) I've obtain 2 notes from grompp
>>
>> NOTE 1 [file ./mdps/em.mdp]:
>>    For energy conservation with switch/shift potentials, rlist should be
>> 0.1
>>    to 0.3 nm larger than rvdw.
>>
>> NOTE 2 [file ./mdps/em.mdp]:
>>    The sum of the two largest charge group radii (0.078024) is larger
>> than
>>    rlist (1.200000) - rvdw (1.200000)
>>
>
> As I recall, there is a small bug where grompp ignores rlistlong when
> printing
> this message.  The simulation will be fine.
>
> -Justin
>
>>
>> Should I increase Rlist to the 1.4 ( as well as rcoulomb to the same
>> value because of PME) ?
>>
>> 2012/12/7, Justin Lemkul <jalemkul at vt.edu>:
>>>
>>>
>>> On 12/7/12 1:19 PM, James Starlight wrote:
>>>> Justin,
>>>>
>>>> following to your advise I've tried to use charmm 27 ff for simulation
>>>> of my protein-cGMP complex ( ligand was parametrized by Swiss Param
>>>> server).
>>>>
>>>> Could you provide me with the cut-offs for vdw as well as
>>>> electrostatics suitable for simulation in charmm27 and 36 force
>>>> fields?
>>>>
>>>
>>> http://lists.gromacs.org/pipermail/gmx-users/2012-September/074717.html
>>>
>>>> Does anybody know another servers for parametrization of the ligands
>>>> for charmm simulation in gromacs?
>>>>
>>>
>>> If Google can't find it, it probably doesn't exist.
>>>
>>> -Justin
>>>
>>>> 2012/12/7, Justin Lemkul <jalemkul at vt.edu>:
>>>>>
>>>>>
>>>>> On 12/7/12 11:42 AM, James Starlight wrote:
>>>>>> Justin,
>>>>>>
>>>>>> ligand-only simulation in vacuum have been finished with the same
>>>>>> errors
>>>>>> :)
>>>>>>
>>>>>> Step 19200, time 38.4 (ps)  LINCS WARNING
>>>>>> relative constraint deviation after LINCS:
>>>>>> rms 0.025443, max 0.140660 (between atoms 1 and 3)
>>>>>> bonds that rotated more than 30 degrees:
>>>>>>     atom 1 atom 2  angle  previous, current, constraint length
>>>>>>          1      2   52.2    0.1033   0.0985      0.1000
>>>>>>          1      3   90.1    0.1168   0.1141      0.1000
>>>>>>
>>>>>> Step 19200, time 38.4 (ps)  LINCS WARNING
>>>>>> relative constraint deviation after LINCS:
>>>>>> rms 0.025376, max 0.140474 (between atoms 1 and 3)
>>>>>> bonds that rotated more than 30 degrees:
>>>>>>     atom 1 atom 2  angle  previous, current, constraint length
>>>>>>          1      2   52.0    0.1033   0.0988      0.1000
>>>>>>          1      3   90.0    0.1168   0.1140      0.1000
>>>>>> step 19200, will finish Sat Dec  8 04:10:49 2012
>>>>>> Step 19201, time 38.402 (ps)  LINCS WARNING
>>>>>> relative constraint deviation after LINCS:
>>>>>> rms 0.052715, max 0.293316 (between atoms 1 and 3)
>>>>>> bonds that rotated more than 30 degrees:
>>>>>>     atom 1 atom 2  angle  previous, current, constraint length
>>>>>>          1      2   49.3    0.0988   0.0993      0.1000
>>>>>>          1      3   90.0    0.1140   0.1293      0.1000
>>>>>>
>>>>>> Step 19201, time 38.402 (ps)  LINCS WARNING
>>>>>> relative constraint deviation after LINCS:
>>>>>> rms 0.052900, max 0.294281 (between atoms 1 and 3)
>>>>>> bonds that rotated more than 30 degrees:
>>>>>>     atom 1 atom 2  angle  previous, current, constraint length
>>>>>>          1      2   49.6    0.0988   0.0989      0.1000
>>>>>>          1      3   90.0    0.1140   0.1294      0.1000
>>>>>>
>>>>>> Step 19202, time 38.404 (ps)  LINCS WARNING
>>>>>> relative constraint deviation after LINCS:
>>>>>> rms 0.035783, max 0.198482 (between atoms 1 and 3)
>>>>>> bonds that rotated more than 30 degrees:
>>>>>>     atom 1 atom 2  angle  previous, current, constraint length
>>>>>>          1      2   33.3    0.0989   0.0984      0.1000
>>>>>>          1      3   89.9    0.1294   0.1198      0.1000
>>>>>>
>>>>>> -------------------------------------------------------
>>>>>> Program mdrun, VERSION 4.5.4
>>>>>> Source code file: /tmp/gromacs-4.5.4/src/mdlib/constr.c, line: 176
>>>>>>
>>>>>> Fatal error:
>>>>>> Too many LINCS warnings (1000)
>>>>>>
>>>>>>
>>>>>> ITs intresting that the same files (like step-1c_n11.pdb) have been
>>>>>> produced already during CG minimisation phase but the system have not
>>>>>> been crashed. So I suppoe that the problem is in geometry of the
>>>>>
>>>>> If you get errors in the minimization, that should indicate that
>>>>> either
>>>>> the
>>>>>
>>>>> topology or configuration is flawed.
>>>>>
>>>>>> mollecule. The direct comparison of the bonds and angles with the rtp
>>>>>> (DGUA) is non trivial again because of differs in the file formats
>>>>>> representation :)
>>>>>>
>>>>>
>>>>> Most troubleshooting is not easy.  You can also create a topology for
>>>>> DGUA
>>>>> from
>>>>> just about any DNA structure and compare its .top against your cGMP
>>>>> topology.
>>>>>
>>>>> -Justin
>>>>>
>>>>> --
>>>>> ========================================
>>>>>
>>>>> Justin A. Lemkul, Ph.D.
>>>>> Research Scientist
>>>>> Department of Biochemistry
>>>>> Virginia Tech
>>>>> Blacksburg, VA
>>>>> jalemkul[at]vt.edu | (540) 231-9080
>>>>> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>>>>>
>>>>> ========================================
>>>>> --
>>>>> gmx-users mailing list    gmx-users at gromacs.org
>>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>>>>> * Please search the archive at
>>>>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>>>>> * Please don't post (un)subscribe requests to the list. Use the
>>>>> www interface or send it to gmx-users-request at gromacs.org.
>>>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>>>
>>>
>>> --
>>> ========================================
>>>
>>> Justin A. Lemkul, Ph.D.
>>> Research Scientist
>>> Department of Biochemistry
>>> Virginia Tech
>>> Blacksburg, VA
>>> jalemkul[at]vt.edu | (540) 231-9080
>>> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>>>
>>> ========================================
>>> --
>>> gmx-users mailing list    gmx-users at gromacs.org
>>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>>> * Please search the archive at
>>> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>>> * Please don't post (un)subscribe requests to the list. Use the
>>> www interface or send it to gmx-users-request at gromacs.org.
>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>
>
> --
> ========================================
>
> Justin A. Lemkul, Ph.D.
> Research Scientist
> Department of Biochemistry
> Virginia Tech
> Blacksburg, VA
> jalemkul[at]vt.edu | (540) 231-9080
> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>
> ========================================
> --
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> * Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>



More information about the gromacs.org_gmx-users mailing list