[gmx-users] qm/mm LA not found

Groenhof, Gerrit ggroenh at gwdg.de
Tue Nov 8 19:08:09 CET 2016 

Hi,

I could reproduce your problem and may know the reason.

I had not realised you were using Gromos force field. Normally one uses the gb_XX defines after each pair of bonded atoms to inform grompp of the values the bond parameters. See for example how it is done in your topol.top. 

Without information (i.e. bond distance and force constant) in the topology file, grompp cannot proceed.

You wrote it worked for 'pure' MM. I suppose that in that .itp file there are defines (gb_XX) or explicitly included values for bond length and force constant?

Best,

Gerrit




That's strange indeed. You showed a copy from the bonded section.

Which was line 133? bond 24-23?

The linkatom is defined as a virtual site and has no mass?

I can only guess, but perhaps if I can see the input files, I may understand quicker what goes wrong here.

Best,

gerrit





Hi Gerrit,

yes, grompp complains only about bond type 1. It gives only error for
lines where the bond types are described, nothing about angle or torsion
types.
I was able to run normal MM, as well as QM/MM if I get the whole
cofactor in QM
region, meaning if all bond types in cofactor are type 5 it runs without
problems.

Greetings
Sylwia


On 07.11.2016 14:50, Groenhof, Gerrit wrote:
> Hi,
>
> Do I understand correctly that the grompp complains about a bond that is still bond type 1?
>
> If so, then the error message suggests that grompp cannot access a ff*bin.itp with the parameters for the atom pair.
>
> Usually you'd also get additional errors for missing angle and torsion types. Do you get those?
>
> Can you run without QM/MM, i.e. a 'normal' MM simulations with this topology?
>
> Gerrit
>
>
> Hi Gerrit,
>
> thank you for your hint. There was no LA entry in the ff*nb.it file. I
> have added it there.
> But now I get a new error. I have a protein with cofactor. Part of the
> cofactor is in QM part
> and the rest in MM. For the cofactor I have a separate *itp file as for
> the rest of protein.
> By preparing the calculation I get and error for all MM atoms:
>
> ERROR 1 [file DRGGMX_groups_v3_QMMM.itp, line 133]:
>     No default Bond types
>
> example from *itp file
>
> [ bonds ]
> ; ai  aj  fu    c0, c1, ...
> ...
>     21  22   5     ;   C9A  N10
>     22  23   5     ;   N10  C1' ;'
>     24  23   1     ;   C2'  C1'
>     24  25   1     ;   C2'  O2'
>
> There is no error for atoms with bond type 5, but for all with bond type 1.
>
> Greetings
> Sylwia
>
>
> On 11/05/2016 03:40 PM, Groenhof, Gerrit wrote:
>> Is the LA also defined for the force field you're using? Can you check whether there is a LA entry in the ff*nb.it of the force field? If not, add it there manually.
>>
>> Best,
>> gerrit
>>
>> Message: 2
>> Date: Fri, 4 Nov 2016 16:10:03 +0100
>> From: Sylwia Kacprzak <sylwia.kacprzak at physchem.uni-freiburg.de>
>> To: gromacs.org_gmx-users at maillist.sys.kth.se
>> Subject: [gmx-users] qm/mm LA not found
>> Message-ID: <581CA4CB.2050002 at physchem.uni-freiburg.de>
>> Content-Type: text/plain; charset=utf-8; format=flowed
>>
>> Dear all,
>>
>> I have set qm/mm calculations with Gromacs 5.1.4 and Orca.
>> I treat first the full cofactor on a QM level. However when I cut the
>> bond to take only part of it for QM and
>> introduce link atom according to the manual with link atom (LA) I get
>> following error.
>>
>> -------------------------------------------------------
>> Program gmx grompp, VERSION 5.1.4
>> Source code file:
>> /home/widu/gromacs-5.1.4/src/gromacs/gmxpreprocess/toppush.c, line: 1345
>>
>> Fatal error:
>> Atomtype LA not found
>>
>> Any ideas?
>>
>> Greetings
>> Sylwia
>>
>>
>>
>
>
> ------------------------------
>
> --
> 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.
>
> End of gromacs.org_gmx-users Digest, Vol 151, Issue 22
> ******************************************************




------------------------------

Message: 5
Date: Mon, 7 Nov 2016 18:06:34 +0000 (UTC)
From: m g <dr_m_ganj at yahoo.com>
To: "gromacs.org_gmx-users at maillist.sys.kth.se"
        <gromacs.org_gmx-users at maillist.sys.kth.se>
Subject: [gmx-users] convergence
Message-ID: <951804911.2088266.1478541994845 at mail.yahoo.com>
Content-Type: text/plain; charset=UTF-8

Dear Justin Lemkul,what must be done for checking of convergence of the umbrella sampling simulations? which kind of parameter must be calculated?
Many thanks

------------------------------

Message: 6
Date: Mon, 7 Nov 2016 23:03:06 +0100
From: Szil?rd P?ll <pall.szilard at gmail.com>
To: Discussion list for GROMACS users <gmx-users at gromacs.org>
Subject: Re: [gmx-users] Can't locate AMD SDK opencl, among else,
        during compile
Message-ID:
        <CANnYEw5DVDjgPWxV9xX3j0UWoLtow7Hho_oZVRGofhKG8KX9AA at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8

Jacob,

Not sure where did things go wrong; please provide full cmake
invocation and it would help to see the contents of CMakeCache.txt
(upload it e.g. to pastebin).

On the other hand, you do no need to link against the SDK
libOpenCL.so, you can link against the generic ICD loader; it's the
ocl-icd-libopencl1 package on Ubuntu (and use opencl-headers for the
headers).

Secondly, the upgrade to Ubuntu 16.04 does complicate things somewhat
as there is no stable AMD driver for that release and some tweaks are
necessary to get best performance. None of that is related to GROMACS,
but to the state of the Linux AMD software stack.

Cheers,
--
Szil?rd


------------------------------

--
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.

End of gromacs.org_gmx-users Digest, Vol 151, Issue 25
******************************************************

------------------------------

--
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.

End of gromacs.org_gmx-users Digest, Vol 151, Issue 27
******************************************************

More information about the gromacs.org_gmx-users mailing list