I tried to make bigger graphite about 200 atoms.
I just found that UCSF chimera can make this structure from the unit
cell pdb(2c.pdb) file by supercell operation and it can also generate
proper bonds between carbons and finally I can see a good graphite
with all 6-membered rings(c10-1.pdb).
I loaded this pdb with previous generated prepi file from 2-carbon
unit cell(2c.prepi).
--------------------------------------------------
source leaprc.ff99SB
source leaprc.gaff
loadamberprep c2.prepi
gr=loadpdb c10-1.pdb
savepdb gr c10amb.pdb
solvatebox gr TIP3PBOX {0 0 20}
saveamberparm gr c10.prmtop c10.inpcrd
---------------------------------------------------
After loading prmtop and inpcrd files, I see additional bonds between
two edge carbons. And after minimization with waters, I see tube like
structure instead of flat surface. I think leap generate wrong bonds.
======================================
sam wat heat 1
&cntrl
imin = 0,
irest = 0,
ntx = 1,
ntb = 1,
cut = 10.0,
ntr = 1,
ntc = 2,
ntf = 2,
tempi = 0.0,
temp0 = 50.0,
ntt = 3,
gamma_ln = 1.0,
nstlim = 10000, dt = 0.0005
ntpr = 50, ntwx = 100, ntwr = 1000
/
Hold the SAM fixed
50.0
RES 1 100
END
END
===================================
And with the slow heating from 0 to 50K, I see NaN error again.
Could you give me any idea how to solve all these problem?
Thank you.
Bongkeun Kim
Quoting Bill Ross <ross.cgl.ucsf.EDU>:
>> I just went back to the residue level.
>>
>> I made a small graphite with 4 ring including 16 atoms(4-1.pdb)
>> And I was able to have prepi and frcmod file after
>> -----------
>> antechamber -fi mol2 -fo prepi -i 4c.mol2 -o 4c.prepi -j both -c bcc -rn GRP
>> parmchk -i 4c.prepi -o 4c.frcmod -f prepc
>> ------------
>
> One thing just occurred to me: I see in your prepi that your
> carbons have different types. E.g.
>
>> 4 C cg M 3 2 1 1.540 111.208 180.000 -0.100300
>> 5 C1 ch M 4 3 2 1.421 30.003 177.258 0.016700
>> 6 C2 cf M 5 4 3 1.421 119.918 175.254 -0.107800
>> 7 C3 c1 M 6 5 4 1.421 119.973 179.997 0.288100
>> 17 C13 ce M 16 15 14 1.421 119.973 -4.747 -0.107800
>
> I expect that and presumably the charges reflect the fact that
> you have parameterized your rings as a standalone molecule. I
> am not sure what antechamber can do better, but it seems like
> you would want more uniformity in a larger sheet. This is why
> I was originally discussing the need to have edge and corner
> residues distinct from the ones that would make up the body of
> the sheet.
>
> Maybe someone else can comment on this and whether antechamber is
> appropriate to the task. Perhaps you could use the current prepi
> and make the atom types all match the inner carbon type - my
> impression is that the sheet is a hexagonal lattice with all
> interior atoms topologically the same.
>
> I'm not sure what the charges would be in such a situation - 0?
> So you may be able to proceed without antechamber or other
> computational parameterization approaches if you don't care
> about edge effects.
>
>> With XtalUtilities, I have a bigger graphite structure 44.pdb by
>> ------------------------------------
>> PropPDB -p 4-1.pdb -o 44.pdb -X 2.461 -Y 2.842 -Z 6.708 -a 90 -b 90 -g
>> 90 -ix 4 -iy 4 -iz 1
>> --------------------------------------------
>>
>> But after running leap, I got many missing parameter error than were
>> not generated parmchk.
>> Could you help me how to fix this problem?
>
> Possibly this is from added bonds involving the edge types as you
> stitch the residues together? It would have been good to see the
> messages. In any case, if you can reasonably change all the types
> to the interior one, these messages might go away.
>
> Bill
>
> _______________________________________________
> 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 Wed Oct 20 2010 - 23:30:04 PDT
