Re: [AMBER] Antechamber atom types and all_nuc94.in atom descriptions

From: FyD <fyd.q4md-forcefieldtools.org>
Date: Fri, 26 Jul 2013 10:36:58 +0200

Dear Kamali,

See http://q4md-forcefieldtools.org/REDS/faq.php#5
When describing the problem you encountered with R.E.D. Server, please
also provide the 'PXXXX' R.E.D. Server job name in the body of your
email so that we can more easily assist you.

> I'm trying to make a library for a central protonated adenosine, and to
> compare charge methods I've been using both option 2 and option 3 with
> single and multiple conformations in the RED Server. I have not been using
> any intra- or inter-molecular charge constraints; I think this is all
> right, but I just wanted to double-check.

I guess option 2 & option 3 corresponds to:
http://q4md-forcefieldtools.org/Tutorial/images/Nuc-frag.jpg
i.e.
   Use R.E.D. IV
   Use R.E.D. IV for auto... nucleotide fragments
right?

> For option 2, I'm uploading central fragments of the nucleotide (with a 5'
> phoshpate and 3'O), so I think no inter-molecular charge constraints need
> to be specified, is that right? And I have not specified any
> intra-molecular constraints but have noticed, for example, that adenine's
> exocyclic amine hydrogens and the nonbridging oxygens have the same
> charges.

For option 2 (if I well understood); see :
http://q4md-forcefieldtools.org/Tutorial/Tutorial-3.php#26
you need to upload whole molecules _and not_ molecular fragments!
   -> R.E.D. generates mol. fragments _from_ the whole molecules
uploaded by the user based on the constraints defined in the P2N
files. See once again:
http://q4md-forcefieldtools.org/Tutorial/Tutorial-3.php#26
First P2N file for dimethylphosphate (DMP):
http://q4md-forcefieldtools.org/Tutorial/P2N/All-frag-Nuc/Mol_red1.p2n
  -> constraints (between the 2 methyl groups of DMP and the HO5 & HO3
groups of the nucleoside) are defined in this P2N file:
    REMARK INTER-MCC 0.0 | 1 2 | 10 11 12 13 | 1 2
    REMARK INTER-MCC 0.0 | 1 2 | 1 2 3 4 | 3 4
Second P2N file:
http://q4md-forcefieldtools.org/Tutorial/P2N/All-frag-Nuc/Mol_red2.p2n
i.e. the nucleoSIDE.

> Similarly, I shouldn't need to specify any intra- or inter-molecular charge
> constraints because I submit the nucleoside and Option 3 supplies the
> dimethylphosphate. I was assuming that Option 3 would also take care of the
> inter-molecular charge constraints for the dimethyl phosphate.

In option 3 (if I well understood again); see :
http://q4md-forcefieldtools.org/Tutorial/Tutorial-3.php#27

you just provide the P2N file for your nucleoside; no need to add the
constraints - they are automatically determined...

regards, Francois



> On Sun, Jul 21, 2013 at 11:01 PM, Kamali Sripathi <ksripath.umich.edu>wrote:
>
>> Dear Francois,
>>
>> Thanks a lot for the information and the answers. I know that a user
>> bringing down a server is very unlikely, but I often worry about the slight
>> possibility. :)
>>
>> Thanks so much again, and have a great night,
>>
>> Kamali
>>
>>
>> On Sat, Jul 20, 2013 at 12:54 PM, FyD <fyd.q4md-forcefieldtools.org>wrote:
>>
>>> Kamali,
>>>
>>> > Thank you for the confirmation. I just submitted jobs to Ante_RED, and
>>> > everything seems to be working correctly. No problems were caused; I was
>>> > only worried that I may have done something to hurt the server :)
>>>
>>> I forgot: the problem you encountered is reported in the FAQ:
>>> http://q4md-forcefieldtools.org/REDS/faq.php#16
>>>
>>> If you could "hurt the server" as a client - this would mean the code
>>> would have a big problem... BTW recently we have corrected a series of
>>> bugs/limitations making the communications within R.E.D. Server far
>>> more efficient; however when the network, the electricity power or the
>>> file system is down - it does not work ;-)
>>>
>>> regards, Francois
>>>
>>>
>>> > On Sat, Jul 20, 2013 at 6:14 AM, FyD <fyd.q4md-forcefieldtools.org>
>>> wrote:
>>> >
>>> >> Dear Kamali,
>>> >>
>>> >> > It turns out that the standard RESP fitting method using the
>>> HF/6-31G*
>>> >> data
>>> >> > set may work for me and my collaborator after all, so we are trying
>>> all
>>> >> > different methods (antechamber using different basis sets and
>>> solvents
>>> >> with
>>> >> > the HF level of theory, as well as multiple options/conformations in
>>> the
>>> >> > RED server) to see how the free energy calculations differ between
>>> the
>>> >> > methods. I've been using the RED server all day, but have just now
>>> had
>>> >> > trouble generating P2N files. I got a message that some sort of
>>> >> maintenance
>>> >> > was going on, and since then the submission screen that I've gotten
>>> after
>>> >> > inputting my PDB files has told me that my file uploads have failed.
>>> I've
>>> >> > attached an image of this screen. Is there maintenance going on, or
>>> is
>>> >> > there something else wrong?
>>> >>
>>> >> Yes - we got a file system problem sometime yesterday in R.E.D.
>>> Server...
>>> >> All should work correctly by now; contact me again if there is a
>>> problem.
>>> >> Sorry for the trouble this problem caused...
>>> >>
>>> >> regards, Francois
>>> >>
>>> >>
>>> >>
>>> >> > On Sat, Jul 13, 2013 at 2:14 PM, Kamali Sripathi <ksripath.umich.edu
>>> >
>>> >> wrote:
>>> >> >
>>> >> >> Dear Francois,
>>> >> >>
>>> >> >> Thank you very much for your detailed reply. I'm attached to
>>> antechamber
>>> >> >> because this protocol is what my collaborators have used
>>> successfully to
>>> >> >> generate libraries. Since you've told me that I can modify the RED
>>> III x
>>> >> >> source code and we have also narrowed down our search in terms of
>>> basis
>>> >> >> set/theory level combinations, I think I will also try RED IIIx as
>>> well.
>>> >> >>
>>> >> >> Thank you so much again and have a great weekend,
>>> >> >>
>>> >> >> Kamali
>>> >> >>
>>> >> >>
>>> >> >> On Wed, Jul 10, 2013 at 4:16 AM, FyD <fyd.q4md-forcefieldtools.org>
>>> >> wrote:
>>> >> >>
>>> >> >>> Dear Kamali,
>>> >> >>>
>>> >> >>> IOp(6/33=2) corresponds to the default algo. of the Kollman's
>>> group.
>>> >> >>> iop(6/33=2,6/41=10,6/42=17) does not correspond to the default
>>> algo...
>>> >> >>> -> I think you should read to understand what these IOP means.
>>> >> >>> http://www.gaussian.com/g_tech/g_ur/k_iop.htm
>>> >> >>> http://www.gaussian.com/g_tech/g_iops/ov6.htm
>>> >> >>> IOp(6/33=2) is not documented; it pnly means 'print the MEP'...
>>> >> >>>
>>> >> >>> Reliable calculations with Antechamber?
>>> >> >>> Do not ask me - I do not use/support Antechamber.
>>> >> >>>
>>> >> >>> You might also be interested in reading the R.E.D. article:
>>> >> >>> http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2918240/
>>> >> >>> Problems (orientation & conformation, algo in MEP computation,
>>> >> >>> chemical equivalencing and charge fitting, etc...) related to MEP
>>> >> >>> computation are described; I hope this describe why R.E.D. was
>>> >> >>> developed...
>>> >> >>>
>>> >> >>> regards, Francois
>>> >> >>>
>>> >> >>>
>>> >> >>> > I've decided to just use the Antechamber method for now, because
>>> I'm
>>> >> >>> more
>>> >> >>> > familiar with it. Initially, what I had done is submit output
>>> from
>>> >> >>> Ante_RED
>>> >> >>> > to Gaussian. I only had a Link 0 header like this:
>>> >> >>> >
>>> >> >>> > *%Mem=4096MB*
>>> >> >>> > *%NProc=8*
>>> >> >>> > *
>>> >> >>> > *
>>> >> >>> > *#P HF/6-31G* SCF=(Conver=6) NoSymm Test*
>>> >> >>> > * Pop=mk IOp(6/33=2) GFInput GFPrint*
>>> >> >>> > *
>>> >> >>> > *
>>> >> >>> > *MEP - Input generated by R.E.D.-IV MOLECULE*
>>> >> >>> >
>>> >> >>> > while changing the basis set and level of theory. Typically, for
>>> >> >>> > antechamber, my input for the Gaussian has Link0 and Link1
>>> before and
>>> >> >>> after
>>> >> >>> > the molecular structures:
>>> >> >>> >
>>> >> >>> > *%NProc=8*
>>> >> >>> > *%mem=4096MB*
>>> >> >>> > *%chk=sa_charged_gauss101.chk*
>>> >> >>> > *#p opt freq HF/6-31G* formcheck*
>>> >> >>> > *
>>> >> >>> > *
>>> >> >>> > *Title Card Required*
>>> >> >>> > *
>>> >> >>> > *
>>> >> >>> > *--- molecular structure ---*
>>> >> >>> > *
>>> >> >>> > *
>>> >> >>> > *--Link1--*
>>> >> >>> > *%chk=sa_charged_gauss101.chk*
>>> >> >>> > *%mem=4096MB*
>>> >> >>> > *%NProc=8*
>>> >> >>> > *#p HF/6-31G* geom=allcheck guess=read pop=mk
>>> >> >>> iop(6/33=2,6/41=10,6/42=17)*
>>> >> >>> >
>>> >> >>> > Could I submit the first through antechamber and still have
>>> reliable
>>> >> >>> > calculations? I know that the grid information is missing in the
>>> >> first
>>> >> >>> > Link0 section I sent you.
>>> >> >>>
>>> >> >>>
>>> >> >>> > On Sun, Jul 7, 2013 at 12:03 PM, FyD <
>>> fyd.q4md-forcefieldtools.org>
>>> >> >>> wrote:
>>> >> >>> >
>>> >> >>> >> Dear Kamali,
>>> >> >>> >>
>>> >> >>> >> > Thank you so much for your detailed reply. I have one
>>> follow-up
>>> >> >>> question
>>> >> >>> >> > for you: Do you know if it is possible to use Antechamber to
>>> >> generate
>>> >> >>> >> RESP
>>> >> >>> >> > charges using a level of theory other than HF/6-31G*?
>>> >> >>> >>
>>> >> >>> >> I would answer try and see ;-) Antechamber (and R.E.D.) just
>>> take
>>> >> the
>>> >> >>> >> MEP computed by QM and perform charge fitting whatever is the QM
>>> >> >>> >> theory level...
>>> >> >>> >>
>>> >> >>> >> You could also download R.E.D. III.x; modify the source code so
>>> that
>>> >> >>> >> the 'RESP-A1' charge model use the QM theory level of your
>>> choice...
>>> >> >>> >>
>>> >> >>> >> regards Francois
>>> >> >>> >>
>>> >> >>> >> PS If you need it we can modify the source of R.E.D. for you...
>>> >> >>> >>
>>> >> >>> >>
>>> >> >>> >> > On Thu, Jul 4, 2013 at 10:51 AM, FyD <
>>> >> fyd.q4md-forcefieldtools.org>
>>> >> >>> >> wrote:
>>> >> >>> >> >
>>> >> >>> >> >> Dear Kamali,
>>> >> >>> >> >>
>>> >> >>> >> >> > I actually do have another question. I would like to
>>> generate
>>> >> >>> charges
>>> >> >>> >> >> using
>>> >> >>> >> >> > a basis set besides HF/6-31G* and AM1-BCC, and I was
>>> wondering
>>> >> if
>>> >> >>> (a)
>>> >> >>> >> the
>>> >> >>> >> >> > RED Server has or will have such a potential, and (b) if
>>> you
>>> >> have
>>> >> >>> any
>>> >> >>> >> >> > suggestions about generating charges with other,
>>> non-standard
>>> >> >>> basis
>>> >> >>> >> sets.
>>> >> >>> >> >>
>>> >> >>> >> >> The use of HF/6-31G* is related to implicit polarization
>>> >> requested
>>> >> >>> in
>>> >> >>> >> >> the additive Amber force field model. See
>>> >> >>> >> >> http://ambermd.org/doc6/html/AMBER-sh-19.4.html#sh-19.4
>>> >> >>> >> >> Be aware of what you do when using MP2 or DFT methods. For
>>> >> instance
>>> >> >>> >> >> MP2/aug-cc-pTVZ is used in the non-additive/polarizable FF
>>> model.
>>> >> >>> >> >>
>>> >> >>> >> >> Concerning AM1-BCC, the later approach is supposed to be an
>>> >> >>> >> >> 'approximation/simplification' of the RESP approach.
>>> Considering
>>> >> >>> that
>>> >> >>> >> >> HF/6-31G* cost nothing (from a cpu point of view) our days I
>>> do
>>> >> not
>>> >> >>> >> >> understand why one would need an 'approximation' of the RESP
>>> >> >>> method...
>>> >> >>> >> >> and obviously approximation/simplification has nothing to do
>>> with
>>> >> >>> >> >> breaking chemical equivalencing (just my personal opinion)...
>>> >> >>> >> >>
>>> >> >>> >> >> R.E.D. Server charge models are described at:
>>> >> >>> >> >> http://q4md-forcefieldtools.org/REDS/popup/popkeyword.php
>>> >> >>> >> >> R.E.D. Server Development charge models are described at:
>>> >> >>> >> >>
>>> >> >>>
>>> http://q4md-forcefieldtools.org/REDS-Development/popup/popkeyword.php
>>> >> >>> >> >>
>>> >> >>> >> >> regards, Francois
>>> >>





_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jul 26 2013 - 02:00:03 PDT
Custom Search