Dear Moitrayee,
This looks like far better. Last comments:
- So far the mol3 file format is mainly useful for molecular
fragments. You have a whole molecule; thus, the mol2 file format is
the best choice (i.e. it can be used as a FF lib in the LEaP program
and easily displayed in a graphical program such as vmd where
topology, atom names, charges values (& atom types) can be visually
checked; and it is directly generated by R.E.D.)...
- If you use leaprc.ff99SB (or any newer version) I do not think so
you need to build a new frcmod file. All the FF parameters for citrate
should be in parm99.dat. Moreover, my feeling is that such an accuracy
for bond & angle FF constants in gaff.dat is useless; just compare the
values in gaff.dat vs parm99.dat, and read the original papers
(parm96, parm98 & parm99).
For instance:
[fyd.master0 parm]$ egrep "^c3-c3 " gaff.dat
c3-c3 303.1 1.5350 SOURCE1 14664 0.0048
[fyd.master0 parm]$ egrep "^CT-CT " parm99.dat
CT-CT 310.0 1.526 JCC,7,(1986),230; AA, SUGARS
-> such an accuracy "303.1" is meaningless: better using parm99.dat...
Same remark for:
[fyd.master0 parm]$ egrep "^c3-h1 " gaff.dat
c3-h1 335.9 1.0930 SOURCE3 2175 0.0082
[fyd.master0 parm]$ egrep "^CT-H1 " parm99.dat
CT-H1 340.0 1.090 changed from 331 bsd on NMA nmodes; AA, RIBOSE
-> such an accuracy "335.9" is meaningless: better using parm99.dat...
- The off file format has the same role than the mol3/mol2 file format(s).
- I would first try to only constraint the HO-OH-CT-H1 dihedral (in
the QM geometry optimization).
- At the end, I would only execute:
xleap -f leaprc.ff99SB
VAR = loadmol2 CIT.mol2
saveamberparm VAR VAR.top VAR.crd
regards, Francois
> Thanks a lot for such a detailed reply and pointers about solving
> this problem.
> I am new to this method and am being much benefited by your comments and
> suggestions.
>
> I have taken care of all the points that you had mentioned. I have added
> hydrogens, I have ensured chemical equivalencing by studying the p2n file in
> pymol, and used two angle constraints. I tried using fewer, but the resultant
> citrate ion was becoming distorted (the CH2 groups were showing
> angles similar
> in that of sp2, so I used constraints as seen in the structure).
>
> Then I used the mol3 format for using in leap, I did loadmol3 and savemol3 in
> xleap followed by a saveoff to generate the CIT.off file. I generate a frcmod
> file using parmchk on the mol2 files. I load the files in leap using
> loadoff and
> saveamberparams command, followed by a saveamberparm to generate
> the prmtop and
> prmcrd files.
>
> Please let me know if I am doing it correctly this time. I am
> attaching the p2n,
> the mol2, the OFF and frcmod files that I am using for the calculation.
>> Dear Moitrayee,
>>
>> Many problems...
>>
>> - All the hydrogens atoms are not present in your molecule; you
>> first need to
> add these (4) hydrogens; you could either use xLEaP or R.E.D. Server:
>> See http://q4md-forcefieldtools.org/REDS/faq.php#20
>> I do not understand how you can end up with atomic charge values in
>> this mol2
> (output) file without all the hydrogen atoms in the P2N input file...
>>
>> - The residue name is not correct in your P2N file; May-be Ante_R.E.D. 2.0
> does not correct this problem; our new R.E.D. version does.
>> Correct that manually; i.e. replace the column '2' by 'CIT' for
>> instance.
>>
>> - Chemical equivalencing is correct in this P2N file; Indeed, here the two
> CH2-COO(-) should be equivalent; this means they should bear the
> same name two
> by two (2 CH2 groups equivalent and 2 COO(-) equivalent). See
> http://q4md-forcefieldtools.org/REDS/popup/popanteredtopequiv.php
> Our new R.E.D.
> version handles that even more automatically; no more P2N files.
>>
>> - The number of geometrical constraints should be kept as small as possible;
> for sure if you constraint all the dihedrals of citrate you will get the same
> structure than the experimental one... please read Cieplak et al.:
>> http://www3.interscience.wiley.com/cgi-bin/abstract/109583237/ABSTRACT
>>
>> - No need to use Antechamber here:
>> - & no need to convert a mol2 file into a prep file; please read:
>> http://q4md-forcefieldtools.org/Tutorial/leap-mol3.php
>> mol3/mol2, off, prep file formats have all the same role: to be a FF library
> file format.
>> - atom types can be easily added in the .mol2 file (simply replace
>> the chemical symbol by the atom types)
>> See http://q4md-forcefieldtools.org/Tutorial/leap-mol2.php
>> & more $AMBERHOME/dat/leap/parm/parm99.dat
>> CT 12.01 0.878 sp3 aliphatic C
>> C 12.01 0.616 sp2 C carbonyl group
>> H1 1.008 0.135 H aliph. bond. to C with 1
>> electrwd. group
>> HO 1.008 0.135 hydroxyl group
>> O2 16.00 0.434 carboxyl and phosphate group oxygen OH
> 16.00 0.465 oxygen in hydroxyl group
>> O2 16.00 0.434 carboxyl and phosphate group oxygen
>>
>> .<TRIPOS>ATOM
>> 1 C -0.898932 -1.107716 -0.639849 CT(**) 1 CIT
> -0.6640(*)
>> 2 C11 0.124175 -0.058604 -0.153210 CT 1 CIT
>> 1.0748 3
> O 0.288277 -0.676295 1.120574 OH 1 CIT -0.6454 4 H
> -0.225218 -0.096106 1.683964 HO 1 CIT 0.3167 5 C10
> 1.354615 -0.039384 -1.034103 CT(**) 1 CIT -0.6640 6 C9
> 2.555128 -0.398100 -0.207257 C 1 CIT 0.8603 7 O8
> 3.121069 0.560707 0.344407 O2 1 CIT -0.8120 8 O7
> 2.917251 -1.585703 -0.258239 O2 1 CIT -0.8120 9 C6
> -0.351104 1.439904 0.084007 C 1 CIT 0.7813
>> 10 O5 -0.304156 2.223542 -0.859272 O2 1 CIT -0.8360
> 11 O4 -0.716263 1.660876 1.251928 O2 1 CIT
> -0.8360 12 C3
> -2.294391 -0.800455 -0.184173 C 1 CIT 0.8603 13 O2
> -2.914347 -0.237493 -1.108215 O2 1 CIT -0.8120 14 O1
> -2.730798 -1.210354 0.899260 O2(!) 1 CIT -0.8120
>>
>> (*) re-derive the atomic charge values with hydrogen atoms...
>> (**) hydrogen atoms of the methylene groups are missing; they
>> should bear the
> H1 atom type...
>> (!) our new R.E.D. version performs atom typing automatically...
>>
>> If I dare (no offense): better understanding what one does instead of using
> programs like black boxes without understanding what one does & why...
>>
>> regards, Francois
>>
>>
>>> I am using angle constraints on the citrate anion (-3) based on the X-ray
> data.
>>> I have attached the P2N file and the .mol2 file. After optimization, I a
> conformation very similar to the experimental conformation. It would
>>> be great if
>>> you take a quick look through them and let me know if I am making some
> obvious/conceptual mistake.
>>> Also I am using antechamber to generate the prep and frcmod file
>>> for the ion
> using the following command:
>>> antechamber -i *.mol2 -fi mol2 -o *.prep -fo prepi -nc -3 -j 5
>>
>>>> Dear Moitrayee,
>>>>> We do have an electron density in the crystal structure to which the
>>>>> citrate was
>>>>> modeled. But I am not sure what sort of experimental data you
>>>>> are referring
> to.
>>>>> It would be immensely helpful if you please elaborate a little
>>>>> more so that
> I
>>> can speak to my experimental collaborator on this issue.
>>>> I would simply check that the structure obtained after geometry
>>>> optimization
> is 'similar' to the experimental one. If not, a
>>>> variation of the
>>> whole molecule approach (described in the previous email) could be to use
> dihedral constraint(s) (a minimum number of constraints) to match the
> experimental conformation.
>>>> You could use R.E.D. Server to perform geometry optimization with dihedral
>>> constraints from a P2N file;
>>>> See http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#CONSTRAINT If
> you do so, please select Gaussian version 2003 because the format of
>>> constraints has changed in Gaussian 09 C.01.
>>>> regards, Francois
>>>>>>> I am performing a MD for which I would require citrate ion (-3)
> parameters. It
>>>>>>> would great if someone could help me with finding if there are
>>>>>>> published
>>> citrate
>>>>>>> ion parameters. I have searched the AMBER archive but did not get them.
>>>>>> Citrate is a complex case; this is small molecule with 3 negative
>>>>>> charges...
>>> If you decide to derive charges for the whole molecule, the
>>>>>> conformation obtained after geometry optimization is unlikely to be
>>> representative of what you want...
>>>>>> If you decide to follow the building block approach (see
>>>>>> http://q4md-forcefieldtools.org/Tutorial/Tutorial-3.php#17) the
> constraints
>>> used during charge derivation are likely to strongly affect the fit...
>>>>>> Do you have experimental data about citrate that could be used to
>>>>>> guide the
>>> charge derivation procedure?
>>>>>> regards, Francois
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Jan 22 2013 - 00:30:03 PST
