Thenk you for the detailed replied
Il 01/11/2011 09:10, FyD ha scritto:
> Dear Francesco,
>
> Sorry I missed your email; when you have questions related to R.E.D.
> Server better send your queries .
> http://lists.q4md-forcefieldtools.org/wws/info/q4md-fft ; there, the
> number of emails is far smaller/should not be missed and mainly
> related to q4md-fft, charge derivation and FF development.
>
>> I have to simulate a protein made by three iron sulphur cluster, namely
>> one 4Fe4S, one "canonical" 3Fe4S and one "not-canonical" 3Fe4S.
>> The topic of the research is the caracterization of the overall protein
>> structure, so I don't need an accurate representation of the clusters.
>> Although I found CHARMM parameters for the 4Fe4S and 3Fe3S clusters, I
>> prefer using AMBER-ILDN force-field. Comparisons with other force-fields
>> show that AMBER-ILDN is the best parameter set to describe protein features.
> Yes, mixing two FF is always a bad idea (there is a web page in the
> Gromacs web site related to this problem). That being said you could
> test if these CHARMM parameters do lead to good results when
> associated with AMBER-ILDN.
Replyng to the same question, David Case states:
"FeS clusters are such unique beasts that there is no "Amber way" or "charmm way" to develop parameters"
So I guess I can use the avalaible CHARMM parameters.
>> So regarding my protein, I guess I can use AMBER for protein. For the
>> iron-sulphur clusters it is enough constraining
>> bonds,angles and dihedrasl among the S and Fe atoms. The problem are
>> atomic charges.
>>
>> I've been running GAMESS-US, both as stand-alone and RED-server, but it
>> fails to converge both with or without geometry optimization.
> When? at the first optimization step?
Yes, it stops at the first stage, but I guess the problem cames from
bad charge-multiplicity couple I gave.
>> I am using the BLYP/6-31+G* model.
> Several comments/questions/suggestions here:
> R.E.D. Server provides GAMESS, Firefly and Gaussian 2003 and 2009.
> - did you try all these programs?
> - did you try R.E.D. Server 'Development'?
> - you can run R.E.D. Server and R.E.D. IV in two modes
> http://q4md-forcefieldtools.org/REDS/popup/popredmodes.php
> this means you can provide a QM geometry optimization output (obtained
> on your workstation) to R.E.D. IV (MEP computation is always carried
> out with NoSymmetry)
> - if you provide us the R.E.D. Server code (PXXX) of your calculations
> that failed we can look at your data and try to help.
> - we also have hidden submission procedures with many new features
> related to transition metal complexes (with a totally new R.E.D. IV
> code). If we think this is needed we can provide these links.
> - finally we can create user defined options within new
> specific/submission procedures; just tell us your needs in this case.
>
>> I've three main questions to address:
>> 1) Has anyone some input file suitable for the RESP-charges calculation
>> with GAMESS-US?
> Sure - R.E.D. interfaces Gaussian/GAMESS/firefly. Charges are highly
> reproducible independently of the QM program interfaced (see
> http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2918240/ for a discussion)
>
> We do use these three QC codes . q4md-forcefieldtools.org; each code
> has its own advantages. Concerning transition metal complexes, all our
> new developments are carried out with the Gaussian 09 program so far.
> We plan to try to extend this new features when using GAMESS and
> Firefly. Did you try using Gaussian 2009?
I submitted a new run using Gaussian 2009 with the RESP-X1 (B3LYP/6-31*)
charge model.
I will report the outcome on the mailing list....
>> 2) Is there any simple rule, or a software, that can help me to
>> calculate iron multiplicity for iron containing molecules?
> For instance using http://www.webelements.com/ Ex: Iron
> - click on Fe
> -> http://www.webelements.com/iron/
> - click on Electron shell properties
> -> http://www.webelements.com/iron/atoms.html
> Ground state electron configuration: [Ar].3d6.4s2 of Iron
> Fe3+ for instance is consequently [Ar].3d5: 5 electrons in 5 d orbitales
> - if 5 single electrons: spin multiplicity = 6 (i.e. I I I I I)
> - if 3 single electrons: spin multiplicity = 4 (i.e. II I I I °)
> - if 1 single electron: spin multiplicity = 2 (i.e. II II I ° °)
> Spin multiplicity is the number of single electron(s) + 1
> See also http://en.wikipedia.org/wiki/Spin_multiplicity
>
> - With a single metal atom this is quite simple.
>
> - With several metal atoms the difficulty is when you want to set
> different spin multiplicities for different metal centers...
>
>> 3) Are parameters for the 3Fe4S avalaible for the AMBER force-field?
> I do not know.
How stated above, It seems that CHARMM parameters can be mixed with
AMBER force-field, at least for these
kinf of molecules
>
>> Any advice doesn't contained in the above questions is well apreciated.
> I am not sure you have used all the options of R.E.D. Server ;-)
I will use more R.E.D. Server option exploring the different features.
In particular, taking care of input geometry, charge and multiplicity.
>
> regards, Francois
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Nov 01 2011 - 03:30:04 PDT
