[gmx-users] Re: Adding New Residue to the Gromos 53.6 ff

Mark Abraham Mark.Abraham at anu.edu.au
Mon Aug 6 10:50:30 CEST 2012 

On 6/08/2012 6:00 PM, James Starlight wrote:
> Mark,
>
>
> The main difficulties is that ATB's output is in the form like
>
>   [ bonds ]
>      1    2    2   0.1000   1.8700e+07
>      1    3    2   0.1000   1.8700e+07
>
>   [ angles ]
> ;  ai   aj   ak  funct   angle     fc
>      2    1    3    2    109.50   380.00
>      2    1    4    2    109.50   425.00
>      3    1    4    2    109.50   425.00
>
>   [ dihedrals ]
> ; GROMOS improper dihedrals
> ;  ai   aj   ak   al  funct   angle     fc
>      7    8   10    9    2      0.00   167.36
>      8   10    9   11    2      0.00   167.36
>     10    9   11    7    2      0.00   167.36
>
> etc
>
> here I must replace the number of atoms by the corresponded real atoms
> but also replace the values for the bonds, angles and dihedrals by the
> Gromacs type codes ( e.g gb_1 etc). This is very routinely for me yet
> :)

You don't need to replace

     1    2    2   0.1000   1.8700e+07


with stuff like

     Cone    Ctwo    gb_234


Stuff like

     Cone    Ctwo    2   0.1000   1.8700e+07

is fine. gb_234 means "paste in the #define in the .itp file"... (seehttp://www.gromacs.org/Documentation/Include_File_Mechanism) so no need to take it out to just paste it in again.  See discussion in manual 5.6.1, which of course you've read ;-)

Most of what you need to do is to map the atom numbers back to atom names. Most of the rest is doing the +N, -C stuff to use the standard peptide linking machinery.

Mark

> I've sent the letter to the ATB's developments with the same
> proposition to five RTP as the output which could be in some cases
> more usefull that ITP.
>
> James
>
> 2012/8/5, Mark Abraham <Mark.Abraham at anu.edu.au>:
>> On 6/08/2012 4:11 PM, James Starlight wrote:
>>> Mark,
>>>
>>>
>>> I suppose that the most trivial way to include such external
>>> topologies is the placement of the ITP file at the bottom of the
>>> topol.top file ( and addition of two covalent bonds with the adjcent
>>> residues in case of GFP chromophore ) like in any potein-ligan system.
>> No, because bonded interactions can only occur within a [moleculetype].
>>
>>> I think that it must be some script foe convertion itp ro rtp because
>>> both of that formats have a lot of differenties so the manually
>>> editing in case of big HET group is very routinely.
>> For the ~25 atoms of the GFP chromophore, given a reliable .itp of
>> Ace-Gly-Cro-Gly-Nme or something, producing an .rtp entry by hand is a
>> job of up to half an hour, if you understand the .rtp format.
>> Particularly if the bonded interactions are from the main forcefield,
>> you've only got to describe atom types, connectivity and charge.
>>
>> You could make a feature request of the ATB people for .rtp output
>> (which might come with supplemental forcefield .itp files).
>>
>> Mark
>>
>>> James
>>>
>>> 2012/8/5, Mark Abraham <Mark.Abraham at anu.edu.au>:
>>>> On 6/08/2012 4:31 AM, James Starlight wrote:
>>>>> Mark,
>>>>>
>>>>> 1) The error was because I've used topology of the CRO generated by
>>>>> the ATB in the ITP form in the RTP file.
>>>> No, the error "Residue 'CRO' not found in residue topology database"
>>>> means you haven't put it in the right place, but until you inspect your
>>>> pdb2gmx output for clues about where it is looking, to help find where
>>>> you haven't followed the right procedure, you're stuck. You've not told
>>>> us your command line or shown us any significant output, so you really
>>>> can't be helped.
>>>>
>>>>> E.g in RTP file I've found that [ atoms ] is simple
>>>>> C     C   0.450     1
>>>>>
>>>>> when in my ITP its more complicated
>>>>> 7     C    1    CRO     C1    5    0.089  12.0110
>>>>>
>>>>> the main difference is that RTP havent atom order where in IPT file
>>>>> the order of each group is specified.
>>>>> Is there any way to convert ITP back to RTP form ?
>>>> No. You'll have to do it by hand.
>>>>
>>>>> 2) I've used specbond.dat to define covalent bonds between C and N
>>>>> termi of chromophore and corresponded termi of the djacent residues (
>>>>> in GFP chromophore is covalently linked in the midle of the alph helix
>>>>> of that protein in both ends)
>>>> Sounds like a recipe for trouble, unless you know how to build a
>>>> two-chain peptide with correct termini into the same [moleculetype], and
>>>> get specbond.dat to work at the same time. It's possible, but why not
>>>> use the built in mechanism for head-to-tail peptide linking?
>>>>
>>>> Mark
>>>>
>>>>> James
>>>>>
>>>>> 2012/8/5, Mark Abraham <Mark.Abraham at anu.edu.au>:
>>>>>> James Starlight wrote
>>>>>>> Dear Gromac's users!
>>>>>>>
>>>>>>> I'm working with the GFP protein which has chromophore (CRO) group
>>>>>>> which is covalently bonded to the protein. I want to add new residue
>>>>>>> to the Gromos 53.6 ff. by means of algorithm described here.
>>>>>>>
>>>>>>>
>>>>>>> For that I've coppied files to my local work dirr and modify all of
>>>>>>> them in accordance to the above tutorial- I've coppied .itp of CRO to
>>>>>>> the aminoacids.rtp
>>>>>>>
>>>>>>> [ CRO ]
>>>>>>> [ atoms ]
>>>>>>> etc
>>>>>>>
>>>>>>> I've defined bonds of the djacent real amino acids in the
>>>>>>> specbond.dat
>>>>>> Seems dangerous from my memory of GFP. You should probably be making a
>>>>>> single residue for the whole chromophore, in which case a model that
>>>>>> does
>>>>>> not need specbond.dat is possible.
>>>>>>
>>>>>>
>>>>>> James Starlight wrote
>>>>>>> Finally I've defined CRO (as the protein) in the residuetypes.dat
>>>>>>>
>>>>>>> When I've try use pdb2gmx I've obtain error
>>>>>>>
>>>>>>> Fatal error:
>>>>>>> Residue 'CRO' not found in residue topology database
>>>>>>>
>>>>>>>
>>>>>>> What I've done wrong?
>>>>>>>
>>>>>> If
>>>>>> http://www.gromacs.org/Documentation/Errors#Residue_'XXX'_not_found_in_residue_topology_database
>>>>>> and its links don't help, you'll have to inspect carefully your output
>>>>>> from
>>>>>> pdb2gmx for clues.
>>>>>>
>>>>>> Mark
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> View this message in context:
>>>>>> http://gromacs.5086.n6.nabble.com/Adding-New-Residue-to-the-Gromos-53-6-ff-tp4999959p4999960.html
>>>>>> Sent from the GROMACS Users Forum mailing list archive at Nabble.com.
>>>>>> --
>>>>>> gmx-users mailing list    gmx-users at gromacs.org
>>>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>>>>>> * Only plain text messages are allowed!
>>>>>> * 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
>>>>>>
>>>> --
>>>> gmx-users mailing list    gmx-users at gromacs.org
>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>>>> * Only plain text messages are allowed!
>>>> * 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
>>>>
>> --
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>> * Only plain text messages are allowed!
>> * 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