Dear Sam,
there is AFAIK a publication describing the RED server with the method
of using different, but standardized molecule orientations in space for
ESP calculation. There exist 'old' publications stating the problem of
the orientation-dependence of ESP-derived atomic charges. This, together
with the possibility to provide your RED server input as supplement to
your work would enhance reproducibility.
The 'validity' would be that the RED server approach tries to compensate
the orientation depencence by using different, but defined orientations
for ESP calculation and then compute the mean. Thus, you'd have 'valid'
ESP-derived atomic charges for a given molecular conformation, a given
charge method, and a given QM compute engine (incl. basis set and
QM-method), defining 'valid' as orientational less dependent.
Howver, you can of course obtain a similar level of reproducibility
doing everything by hand and locally, which may be more comfortable in
some cases - I mentioned the external dependency. Additionally, I'm not
sure, how well RED server can deal with the systems you are interested
in, i..e. organometallic species. Since the server also provides
automatic atom type recognition and makes suggestions for the (missing)
force field parameters, this could be helpful in one case, and
distracting in the other.
Thus, the decision is still up to you.
But I hope, my comments helped a bit.
Best regards,
Anselm
On 12/17/2020 08:37 PM, Sam Walsworth (Researcher) wrote:
> That is a good way summing up pros & cons of Antechamber vs RED Server.
>
> Could you put forward an argument to suggest that the RED server method is better for generating publication quality information since it came from source that is more reproducable and somewhat 'valid'?
>
> Thank you for your insight
> Sam
>
> ________________________________
> From: Dr. Anselm Horn <anselm.horn.fau.de>
> Sent: 17 December 2020 15:56
> To: amber.ambermd.org <amber.ambermd.org>
> Subject: Re: [AMBER] Antechamber vs R.E.D. Server
>
> Dear Sam,
>
> IMHO there is no "better alternative", since atomic charges are per se a
> construct, and RESP charges may be generated in multiple ways.
>
> The advantage of using a local Games-US version together with
> antechamber may be that you have everything locally, full/direct
> control, and no external dependencies.
>
> The advantage of using RED server may be that your charges gain
> reproducibility and some sort of validity: Since RESP charges are known
> to depend on the molecule's orientation in space, the default algorithm
> there uses two "standard orientations" of the molecule (i.e. a rotation
> in space) and computes the mean RESP charges. You can increase the
> number of these fixed orientations, and also use different conformations
> for your molecule - all that can also implemented manually, but may be
> tedious.
>
> Best regards,
>
> Anselm
>
>
>
> On 12/16/2020 08:30 PM, Sam Walsworth (Researcher) wrote:
>> Hi all
>>
>> I'm going to use Antechamber (w/ GAMESS-US) to generate ESP charges and fit them via RESP. I was considering if using R.E.D. Server is a better alternative to GAMESS-US/Antechamber? I will be using organometallic ligands going forward and read that software for generating organometallic parameters such as VFFDT are fully compatible with antechamber, hence why I've committed to the GAMESS-US/Antechamber method thus far.
>>
>> Thanks in advance
>> Sam
>> University of Huddersfield inspiring global professionals.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Dec 18 2020 - 02:00:03 PST
