Dear Jason, Asmita,
> The charges from all_prot_nucleic10.lib come from RESP calculations using
> HF 6-31G*, if I recall correctly. Inter-molecular constraints were used to
> keep the charge of the sugar ring equal to the corresponding base with the
> normal protonation state, so only the charge on the bases changed.
Thanks Jason, but this is rather vague, and for sure charge values
cannot be reproduced/compared with these pieces of information... So
Asmita, the charge differences you report seems pretty good in these
conditions.
regards, Francois
> On Wed, Jul 4, 2012 at 4:16 AM, FyD <fyd.q4md-forcefieldtools.org> wrote:
>
>> Dear Asmita,
>>
>> I do not know where come from the charge values available in
>> all_prot_nucleic10.lib, and I cannot answer instead of the Amber
>> developers... You should send this type of query in the Amber mailing list,
>> and not to my personal email address.
>>
>> That being said a related problem (I guess) has been discussed many times
>> in the Amber mailing list, and a summary of the causes of this problem is
>> available (I guess) in the R.E.D. tools article.
>> See
>> http://www.ncbi.nlm.nih.gov/**pmc/articles/PMC2918240/<http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2918240/>
>> See
>> http://www.ncbi.nlm.nih.gov/**pmc/articles/PMC2918240/#**S7title<http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2918240/#S7title>
>> See the section: "Charge reproducibility"
>>
>> To try to make the answer short, conformation 'greatly' (not defined)
>> affect charge values; and molecular orientation '' affect charge values
>> (all that is rather vague...). This means that even if a well defined
>> conformation is involved in charge derivation two users are likely to
>> generate two different sets of charge values when using the GAMESS, Firefly
>> and Gaussian programs ... or even using only Gaussian. Thus, I would answer
>> that the charge values in all_prot_nucleic10.lib (but once again I have no
>> idea where they come from) are not reproducible simply because one has no
>> information (i) about the conformation(s) involved in charge derivation,
>> and (ii) the molecular orientation(s) is(are) unknown as well... The type
>> of grid of points/surface(s) involved in MEP computation has also to be
>> rigorously defined.
>>
>> This is why R.E.D., R.E.DD.B. and by now R.E.D. Server have been
>> developed. By rigorously controlling all the parameters that affect
>> non-polarizable MEP-based charge values in R.E.D. and by reporting these
>> parameters in R.E.DD.B. the charge values are reproducible by other users.
>> Thus, by using R.E.D. you will not be able to reproduce the charge values
>> available in all_prot_nucleic10.lib (my guess), but other users will be
>> able to reproduce the data you generate.
>>
>> As a test you can take data from R.E.DD.B. and try to reproduce the
>> corresponding charge values; then, apply the same strategy to reproduce
>> charge values in all_prot_nucleic10.lib.
>>
>> I hope this helps.
>>
>> regards, Francois
>>
>>
>> I am having a confusion here regarding library file for protonated nucleic
>>> acid bases. What is the QM scheme used in Amber11/12 to generate partial
>>> charge values for protonated bases in all_prot_nucleic10.lib file given
>>> with amber12.
>>>
>>> To say it more clearly, why partial charge values for an N3 protonated
>>> cytosine using Gaussian09 and RED server(generated by me) are different
>>> from that given in all_prot_nucleic10.lib file. The 2 residues are same.
>>> I
>>> have used HF/6-31G for Gaussian geometry optimization followed by RED
>>> calculation.
>>>
>>> Does that mean that Amber uses a different scheme for calculation of these
>>> values. Also, these files have been added in the recent versions of amber.
>>> They were not present in Amber9. I am in a fix here as to which values to
>>> be used for further MD simulation.
>>>
>>> Here i am giving partial charge values in the two cases:
>>>
>>> N3-protonated cytosine :
>>>
>>> Amber12 values RED values
>>>
>>> N1: 0.1954 0.0520
>>> C2: 0.5039 0.5302
>>> O2: -0.4753 -0.4928
>>> N3: -0.4871 -0.3079
>>> H3: 0.4128 0.3422
>>> C4: 0.6466 0.5653
>>> N4: -0.8363 -0.8737
>>> H41: 0.4518 0.4707
>>> H42: 0.4518 0.4707
>>> C5: -0.4218 -0.3638
>>> C6: 0.0028 -0.0362
>>>
>>> Can you please clarify this point.
>>>
>>> Regards
>>> Asmita
>>>
>>> On Thu, Jun 28, 2012 at 12:27 PM, FyD <fyd.q4md-forcefieldtools.org>
>>> wrote:
>>>
>>> Dear Asmita,
>>>>
>>>> > There is a file called all_prot_nucleic10.lib which is provided in
>>>> > /amber12/dat/leap/lib folder. The file actually has partial charge
>>>> values
>>>> > for protonated nucleic acid bases. My doubt is whether this library
>>>> could
>>>> > be used with recent ff12SB force field for a structure which has both
>>>> > standard and protonated nucleic acid bases. And if we can use it, why
>>>> do
>>>> we
>>>> > need to do a separate Gaussian geometry optimization followed by MEP
>>>> > calculation for protonated bases.
>>>>
>>>> if this FF library file contains what you need Yes you can use/test it
>>>> with the FF of your choice.
>>>>
>>>> in general one derives charges and build FF library(ies) only for new
>>>> molecule(s)/fragment(s).
>>>>
>>>> regards, Francois
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jul 06 2012 - 23:30:03 PDT
