[gmx-users] Build oligosaccharide topology with pdb2gmx

[Elena Lilkova]

Hi,

thanks for the suggestion, Tom. I added one more or condition in the if
statement of the find_nc_ter function in pdb2gmx.c and now it patches the
termini correctly.
I have one more question though. The pdb2gmx output does not state clearly
if it applied the specbonds or not (it only says "10 out of 10 lines of
specbond.dat converted successfully" and this is done before choosing the
termini). The bonds between the different chains that I specified in the
specbods.dat are not present in the topology. How do I know if the chains
(i.e. residues) are linked correctly?

Thanks again,

Elena


> Hi,
>
> You can get around this but slightly modifying the pdb2gmx source code
> (pdb2gmx.c) so that it additionally recognises something like 'sugar'
> and also defining your residues as this in residuetypes.dat. This way
> pdb2gmx should now patch your termini correctly. I have done this in the
> past for oligosaccharides (GROMOS not CHARMM ones though).
>
> Cheers
>
> Tom
>
> On 04/03/15 17:34, Elena Lilkova wrote:
>> Hi,
>>
>> unfortunately, my idea does not work. When I define the sugars in the
>> residuetypes.dat as "Other" it does not patch the termini, and pdb2gmx
>> says:
>>
>> Warning: Starting residue SGN1 in chain not identified as
>> Protein/RNA/DNA.
>> Problem with chain definition, or missing terminal residues.
>> This chain does not appear to contain a recognized chain molecule.
>> If this is incorrect, you can edit residuetypes.dat to modify the
>> behavior.
>>
>> If I declare them to be protein, it tries to patch them, but complains
>> that:
>>
>> Atom O in residue SGN 1 was not found in rtp entry SGN with 28 atoms
>> while sorting atoms.
>>
>> So, I'll just do the big single rtp entry for the whole thing.
>>
>> Thanks anyway and have a nice evening,
>>
>> Elena
>>
>>>
>>> On 3/4/15 9:03 AM, Elena Lilkova wrote:
>>>> Hi, Justin,
>>>>
>>>> thanks for the quick reply.
>>>> I was thinking of making each saccharide a separate chain, so that
>>>> each
>>>> saccharide goes through the termini patching. As I understand it, my
>>>> new
>>>> .tdb should be something like:
>>>>
>>>> ; Link at C4 of res (i-1) for (i)1->4(i-1) linkage
>>>> [ LINK-C4 ]
>>>> [ replace ]
>>>> C4 C4 CC3161 12.011   0.09
>>>> O4 O4 OC301  15.9994 -0.36
>>>> [ delete ]
>>>> HO4
>>>>
>>>> ; Link at C1 of res i for (i)1->4(i-1) linkage
>>>> [ LINK-C1 ]
>>>> [ replace ]
>>>> C1 C1 CC3162 12.011   0.29
>>>> [ delete ]
>>>> HO1
>>>> 2O1
>>>>
>>>> And then adding the following to the specbond.dat:
>>>> RES1   O4   1   RES2   C1    1    0.14     RES1   RES2.
>>>>
>>> Worth a shot.  Please report back if this works; it would be very
>>> useful
>>> to know
>>> and would be a nice benefit for the community.
>>>
>>> -Justin
>>>
>>>> Anyway, eventually I will probably just create one ginormous "residue"
>>>> for
>>>> the whole ilogosacchiride.
>>>>
>>>> Thanky you once again,
>>>>
>>>> Elena
>>>>
>>>>>
>>>>> On 3/4/15 7:54 AM, Elena Lilkova wrote:
>>>>>> Dear GMX community,
>>>>>>
>>>>>> I would like to use the CHARMM 36 carbohydrate force field to
>>>>>> simulate
>>>>>> an
>>>>>> oligosaccharide. I read the pdb2gmx section of the manual, but I am
>>>>>> still
>>>>>> wondering how to build the topology for my oligosaccharide.
>>>>>>
>>>>>> My first question is can I add in the "residuetypes.dat" residues
>>>>>> that
>>>>>> are
>>>>>> neither protein, nor DNA, RNA, water, or ions, but say type sugar.
>>>>>>
>>>>> Listing it simply as "Other" is the straightforward solution.
>>>>>
>>>>>> Next, how can I link the individual saccharides. I would like to use
>>>>>> two
>>>>>> of the several linkages, that are parameterized in the force field
>>>>>> (i.e.
>>>>>> PRES 14aa and PRES 14ab). This means that in both of the two
>>>>>> consecutive
>>>>>> saccharides some atoms should be deleted and some partial charges
>>>>>> should
>>>>>> be changed.
>>>>>> I now how to do this with the psfgen plugin of vmd, but how am I
>>>>>> supposed
>>>>>> to achieve this with pdb2gmx. I should probably make a termini
>>>>>> database
>>>>>> file and a special bonds file, right?
>>>>>>
>>>>> No, the .tdb will only affect the terminal residues and won't do
>>>>> internal
>>>>> patching.  At present, there's no built-in way for pdb2gmx to do
>>>>> patching
>>>>> like
>>>>> CHARMM does.
>>>>>
>>>>> The possible solutions are:
>>>>>
>>>>> 1. Write a script to convert the PSF to GROMACS .top format
>>>>> (eventually
>>>>> we
>>>>> will
>>>>> support PSF natively)
>>>>> 2. Create an .rtp entry for the fully built oligosaccharide, with all
>>>>> of
>>>>> the
>>>>> changes to atom types and charges already made.  The oligo
>>>>> effictively
>>>>> becomes
>>>>> on single "residue."
>>>>>
>>>>> -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
>>>>>
>>>>> ==================================================
>>>>>
>>> --
>>> ==================================================
>>>
>>> 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
>>>
>>> ==================================================
>>>
>>
>>
>
> --
> Dr Thomas Piggot
> University of Southampton, UK.
>
>