Barbara,
a few comments on your long e-mail:
The general setup of the ligand incl. geometry optimization seems fine
to me (I did not check the actual structures, though).
It is unclear to me, why you chose bcc atomic charges, when you have
Gaussian available; you could then also have created RESP charges.
The definition of the mainchain file seems not correct to me: The main
chain defines the backbone atoms, in your case this would be just
C,CA,N, and not the complete series of heavy atoms; in the image you
provided KCX is a modified Lys residue, which is connected to the rest
of the protein via its normal peptide bonds.
Probably, the definition in the mc file and the subsequent backbone
definition in the prep file causes the circular structure you observed.
You may want to have a personal look at the automatically derived atom
types of your system, since these are the basis for the parameter
combinations: If you have a system with ff14SB atom types only, you do
not need to run parmchk2 for gaff parameters.
After your parmchk2 runs, you simply strip out the problematic entries
(ATTN), but you did not seem to revise them, which might be problematic.
Since you did not provide the prep and frcmod files for your ligand,
it's hard to give any further advise.
I'd suggest to restart with a corrected mc file, closely inspect the
prep and frcmod files, and use the parameter files to setup a simple
test system via leap like
mini=sequence {ACE KCX NME}
and work on from that (minimization, small md; if all goes well, then
incorporation into the real protein system).
Depending on the 'importance' of your residue, you might want to create
RESP atomic charges or even use two conformations (alpha/beta) for RESP
charge generation.
(For more specific questions, you could also contact me off the list.)
Hope that helps.
Best,
Anselm
Bioinformatik | NHR.FAU
Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU)
Germany
On 01/05/2023 06:35 AM, Barbara Herlah via AMBER wrote:
> Hello Dear fellow amber users,
>
> I have a question regarding preparation of a non-standard KCX residue that has a charge -1. I helped myself using two tutorials https://ambermd.org/tutorials/advanced/tutorial1/section2.htm and http://ambermd.org/tutorials/basic/tutorial5/
>
> I began my journey with drawing the KCX residue with a charge -1 and ACE and NME caps on each side of the residue there the protien continues. I performed a MMFF94 geometry optimization to get a 3D structure, saved it as -mol2 file and used OpenBabel to convert it to .xyz file. Then I performed a ligand optimization and electrostatic grid calculation of the ligand with Gaussian 16.
>
> Optimization:
>
> %NProcShared=8
> %chk=opt_geom.chk
> %mem=4000MB
> #p opt HF/6-31G(d) nosymm iop(6/7=3) gfinput
>
> Ligand optimization
>
> -1 1
> N 0.91855 -0.24842 -0.05187
> C 0.22768 0.85908 -0.70521
> C -0.66191 0.32043 -1.85691
> O -1.13106 -0.81503 -1.86175
> C 2.77243 -1.75052 0.36762
> C -0.60772 1.66221 0.30755
> C 0.27282 2.31308 1.38789
> C -0.53354 3.15093 2.38690
> C 0.38019 3.78389 3.44261
> N -0.33255 4.59174 4.39467
> C -0.93530 4.07714 5.44751
> O -0.78185 2.85174 5.67876
> O -1.59215 4.94164 6.08269
> N -0.91449 1.25855 -2.84175
> C -1.71657 0.98174 -4.00873
> C 2.19197 -0.60772 -0.42101
> O 2.83372 -0.03308 -1.29552
> H 0.34013 -0.93422 0.41863
> H 0.97890 1.52039 -1.15197
> H -1.35011 1.00836 0.78408
> H -1.17052 2.44232 -0.21929
> H 1.02449 2.94890 0.90407
> H 0.81630 1.53592 1.93842
> H -1.29121 2.52443 2.87118
> H -1.07206 3.94408 1.85340
> H 0.98211 3.03203 3.96609
> H 1.08745 4.45825 2.94564
> H -0.51215 5.56228 4.20415
> H -0.44067 2.15193 -2.79430
> H -2.58875 1.64031 -3.98586
> H -1.12232 1.21573 -4.89488
> H -2.04191 -0.06071 -4.05186
> H 2.98814 -2.58311 -0.30728
> H 3.69884 -1.41310 0.84197
> H 2.09284 -2.09440 1.15132
>
> Using a command g16 < /filename.com/ > /filename.out/
>
>
> El. grid calculation:
>
> %NProcShared=8
> %chk=opt_geom.chk
> %mem=600MB
> #p HF/6-31G(d) nosymm iop(6/33=2,6/41=10,6/42=17) pop=mk guess=read gfinput geom=check
>
> Ligand electrostatic grid calculation
>
> -1 1
>
>
> Using a command g16 < /el_grid.com/ > /el_grid.out/
> Next, I used antechamber:
>
>> antechamber -i el_grid.out -fi gout -bk KCX -fo ac -o KCX.ac -c bcc -at amber -nc -1 -m 1 -rn KCX
>
> Then I created a file KCX_1.mc
> I labeled HEAL_NAME and TAIL_NAME, MAIN_CHAIN of all of the heavy atoms that KCX consists of, OMIT_NAME contains the atoms of ACE and NME caps, PRE_HEAD_TYPE, POST_TAIL_TYPE and CHARGE. See below:
>
>
> HEAD_NAME N1
> TAIL_NAME C2
> MAIN_CHAIN C1
> MAIN_CHAIN O1
> MAIN_CHAIN C4
> MAIN_CHAIN C5
> MAIN_CHAIN C6
> MAIN_CHAIN C7
> MAIN_CHAIN N2
> MAIN_CHAIN C8
> MAIN_CHAIN O2
> MAIN_CHAIN O3
> OMIT_NAME C10
> OMIT_NAME O4
> OMIT_NAME C3
> OMIT_NAME H16
> OMIT_NAME H17
> OMIT_NAME H18
> OMIT_NAME N3
> OMIT_NAME H12
> OMIT_NAME C9
> OMIT_NAME H13
> OMIT_NAME H14
> OMIT_NAME H15
> PRE_HEAD_TYPE C
> POST_TAIL_TYPE N
> CHARGE -1.00
>
> Now I run the script using prepgen:
>> prepgen -i KCX.ac -o KCX_1.prepin -m KCX_1.mc -rn KCX
>
>
>
> Everything is fine and I get no error so far.
>
> Then I continue with parmchk2 and create two frcmod files one with amber and one with gaff parameters as tutorial 5 recomends:
>> parmchk2 -i KCX_1.prepin -f prepi -o KCX_1.frcmod -a Y -p $AMBERHOME/dat/leap/parm/parm10.dat
>
>> grep -v "ATTN" KCX_1.frcmod > KCX_2.frcmod # Strip out ATTN lines
>
>> parmchk2 -i KCX_1.prepin -f prepi -o KCX_3.frcmod
>
> I also checked the NEWPDB.PDB file and it looks good - like a KCX residue should. There are two double bonds between C and O atoms missing in this representation however.
> [Inline image OWAPstImg969435]
>
> Then I prepare leap.in file as follows:
> source leaprc.gaff2
> source leaprc.protein.ff14SB
> source leaprc.water.tip3p
>
> set default PBRadii mbondi3
>
> loadamberprep KCX_1.prepin
> loadamberparams KCX_3.frcmod
> loadamberparams KCX_2.frcmod
>
> a = loadPDB 2jff_OK2.pdb
>
> addIons a Na+ 0
> addIons a Cl- 0
>
> solvateBox a TIP3PBOX 10 iso
>
> savepdb a system_leap_3.pdb
>
> saveamberparm a system_3.top system_3.coord
>
> quit
>
> And I run it using:
>> tleap -f leap.in
>
> Results:
> I get ZERO errors, however I get 10 warnings (more in leap1.log file).
>
> /apps/amber/amber20/bin/teLeap: Warning!
> addIons: 1st Ion & target unit have charges of the same sign:
> unit charge = -2e-06; ion1 charge = -1;
> can't neutralize.
>
> /apps/amber/amber20/bin/teLeap: Warning!
> Converting N-terminal residue name to PDB format: NALA -> ALA
>
> /apps/amber/amber20/bin/teLeap: Warning!
> Converting C-terminal residue name to PDB format: CARG -> ARG
>
> /apps/amber/amber20/bin/teLeap: Warning!
> Converting N-terminal residue name to PDB format: NARG -> ARG
>
> /apps/amber/amber20/bin/teLeap: Warning!
> Converting C-terminal residue name to PDB format: CGLN -> GLN
>
> /apps/amber/amber20/bin/teLeap: Warning!
> Converting N-terminal residue name to PDB format: NGLU -> GLU
>
> /apps/amber/amber20/bin/teLeap: Warning!
> Converting C-terminal residue name to PDB format: CHID -> HID
>>
>> saveamberparm a system_3.top system_3.coord
> Checking Unit.
>
> /apps/amber/amber20/bin/teLeap: Warning!
> There is a bond of 8.716 angstroms between O3 and C2 atoms:
> ------- .R<KCX 198>.A<O3 22> and .R<KCX 198>.A<C2 23>
>
> /apps/amber/amber20/bin/teLeap: Warning!
> There is a bond of 3.202 angstroms between O1 and C4 atoms:
> ------- .R<KCX 198>.A<O1 5> and .R<KCX 198>.A<C4 6>
>
> Building topology.
> Building atom parameters.
> Building bond parameters.
> Building angle parameters.
> Building proper torsion parameters.
> Building improper torsion parameters.
> ** Warning: No sp2 improper torsion term for O2-CT-C-N
> atoms are: O3 C1 C2 N
> old PREP-specified impropers:
> <KCX 198>: -M C1 N1 H1
> <KCX 198>: C8 C7 N2 H11
> <KCX 198>: N2 O2 C8 O3
> <KCX 198>: C1 +M C2 O1
> total 1274 improper torsions applied
> 4 improper torsions in old prep form
>
> When I look the .pdb created by the leap.in, it doesn0t look like the KCX residue is connected to the protein since we can see the dotted line where a residue should be in a cartoon vizualization.
>
> [Inline image OWAPstImg263658]
>
> I additionally created a restart file.
> When I run minization nvt.in script containing:
>
> n water nvt restraint
> &cntrl
> imin = 0, irest =0, ntx = 1, ntxo = 1, ioutfm = 1,
> ntb = 1, pres0 = 1.0, ntp = 0,
> taup = 2.0,
> cut = 10, ntr = 1,
> ntc = 2, ntf = 2,
> tempi = 0.0, temp0 = 50.0,
> ntt = 3, ig = -1, gamma_ln = 1.0,
> nstlim = 10000, dt = 0.0005,
> ntpr = 1, ntwx = 5000, ntwr = 5000,
> vlimit=20
> /
> Hold the protein fixed
> 100.0
> RES 1 433
> END
> END
>
> When I vizualize the first .nc file created it somehow forms wierd bonds and connects the residue in a loop.
> [Inline image OWAPstImg604716]
>
> Now I'm wondering where have I failed with the creation of non-standard residue. And now I have two questions:
>
>
> 1. Regarding the circle that the residue created
>
> Is it the .mc file? How should it be written differently? Is it something else completely?
>
> 2. Regarding the lack of bond created
> Should I have written in leap.in file for the bond to be created?
> bond a./number_of_residue1/.C a./number_of_residue2/.N
> bond a./number_of_residue2/.N a./number_of_residue3/.C
>
> Please advise me, I no longer see what else could I have done differently. Thank you
>
> Barbara
>
>
>
>
>
>
>
>
> Barbara Herlah
>
> Doktorska ¹tudentka / Young Researcher
>
>
> [cid:image001.png.01D69D88.0294FD40]
>
>
> [cid:image002.png.01D69D88.0294FD40]
>
>
>
> Teoretièni odsek / Theory Department
>
> Kemijski in¹titut / National Institute of Chemistry
>
> Hajdrihova 19
>
> SI-1000 Ljubljana
>
> Slovenija / Slovenia
>
> 00386 1 47 60 394
>
> barbara.herlah.ki.si www.ki.si<https://www.ki.si/>
>
>
>
>
>
>
> _______________________________________________
> 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 Thu Jan 05 2023 - 05:30:03 PST
