VMD-L Mailing List

From: Chitrak Gupta (chgupta_at_mix.wvu.edu)
Date: Thu Sep 10 2015 - 12:39:01 CDT

Hello!

I re-did the protocol from start. However, this time when I was optimizing
the charges, I loaded all the Gaussian *.log files from water optimization
and hit "Run optimization", I encountered the following error.

Application halting due to following errors:

Found inappropriate upper bound.

This did not happen the first time. Do you know why this could be happening?

Regards,
Chitrak.

On Wed, Sep 9, 2015 at 10:12 AM, Mayne, Christopher G <cmayne2_at_illinois.edu>
wrote:

> Chitrak,
>
> Yes, you will have to go back and redo the charge optimization step. The
> +0.09 assigned for each hydrogen has to be accounted for when distributing
> charges during the optimization such that the net charge sums to an integer
> (0 for neutral species).
>
> Regards,
> Christopher Mayne
>
> On Sep 9, 2015, at 8:17 AM, Chitrak Gupta wrote:
>
> Hi Christopher,
>
> Thanks a lot for your continued help.
>
> I just realized that when I was creating the original PSF file for my
> residue, I did not change the charges of the aliphatic hydrogens to 0.0900
> (every atom was set to 0). I completed the charge optimization (and removed
> the aliphatic hydrogens as recommended), and that is when I realized that
> those atoms still have a 0.0000 charge.
>
> Do I have to re-do the entire protocol with the correct aliphatic H
> charges? Or can I simply manually change it in the PSF and proceed normally
> from this point?
>
>
>
> Regards,
> Chitrak.
>
> On Tue, Sep 8, 2015 at 4:10 PM, Mayne, Christopher G <cmayne2_at_illinois.edu
> > wrote:
>
>> The guessing algorithm gives you a decent place to start from; hopefully
>> you would need to remove fewer entries than you would have had to add, and
>> thus, saving you time.
>>
>> The water molecules aren't actually bonded to your molecule. When
>> loading in files that don't contain explicit bonding information, e.g.,
>> PDB, VMD employs heuristics to guess where the bonds are based on
>> atom-to-atom distances; once the bonds are defined, they apply to all
>> frames. Since the water is close to your molecule in the first frame of
>> the optimization, VMD thinks there's a bond and applies is to all
>> subsequent frames.
>>
>> Regards,
>> Christopher Mayne
>>
>>
>>
>> On Sep 8, 2015, at 2:57 PM, Chitrak Gupta wrote:
>>
>> Hi Christopher,
>>
>> Thanks for directing me to this link.
>>
>> So from what I understand, for anything but really small molecules, it is
>> a good idea to use your chemical intuition to select donor and acceptor
>> atoms, is that correct?
>>
>> I re-did this step, this time manually selecting the donors and
>> acceptors. However, I still have a few waters moving away from the molecule
>> but are still bonded to it, resulting in unrealistic bond lengths.
>>
>>
>> Regards,
>> Chitrak.
>>
>> On Tue, Sep 8, 2015 at 1:48 PM, Mayne, Christopher G <
>> cmayne2_at_illinois.edu> wrote:
>>
>>> Chitrak,
>>>
>>> This topic has been discussed very recently:
>>>
>>> http://www.ks.uiuc.edu/Research/vmd/mailing_list/vmd-l/26080.html
>>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.ks.uiuc.edu_Research_vmd_mailing-5Flist_vmd-2Dl_26080.html&d=AwMFAg&c=8hUWFZcy2Z-Za5rBPlktOQ&r=wF2NhQZUqE_YMsbjgIM_pn5EzgUuDU4489TyIrVxZMA&m=nxdnwoVkBAlKo6C5YLtAkURGX7u5dduUPtsDXBZtARg&s=pB6bMmybO11TVe1NjdXJtzcxEWwX7jpLnGeR6UteoQk&e=>
>>>
>>>
>>> Regards,
>>> Christopher Mayne
>>>
>>> On Sep 8, 2015, at 10:54 AM, Chitrak Gupta wrote:
>>>
>>> Hi! I am new to the FFTK plugin. I am following the steps of the
>>> tutorial but having problems at the water interaction stage. I used the
>>> "Auto detect indices" to select the hydrogen bonding atoms. It included
>>> even Hydrogens bonded to aliphatic C, and atoms like amide C, carboxyl C,
>>> etc. There were also atoms that were included both in lists of donors and
>>> acceptors. Is that normal?
>>>
>>> After I ran the Gaussian calculations, some of the waters move away from
>>> the molecule. Some other waters also move away but are still bonded,
>>> resulting in severely elongate bonds. Does anyone know how to fix these?
>>>
>>>
>>> Regards,
>>> Chitrak.
>>>
>>>
>>>
>>
>>
>
>