CHARMM Development Project Forums User Discussion & Questions Parameter Set Discussion question generating forcefield parameters

Dear all,

I generated charmm forcefield parameters for multiple engineered residues in a cyclic peptide. When reading the pdb file using charmm, the coordinates were mostly 9999.00 as shown below.

179 11 MVA C1 9999.0000000000 9999.0000000000 9999.0000000000 LIG 11 0.0000000000
180 11 MVA C2 9999.0000000000 9999.0000000000 9999.0000000000 LIG 11 0.0000000000
181 11 MVA O1 9999.0000000000 9999.0000000000 9999.0000000000 LIG 11 0.0000000000
182 11 MVA C6 9999.0000000000 9999.0000000000 9999.0000000000 LIG 11 0.0000000000

I got the error message in the output file like the following:

CHARMM> GENERATE LIG SETUP
NO PATCHING WILL BE DONE ON THE FIRST RESIDUE
NO PATCHING WILL BE DONE ON THE LAST RESIDUE
: No angle parameters for 12 ( CG2O CG31 CG31)
: No angle parameters for 66 ( CG32 CG31 NG32)

In fact, those angles formed only when two neighboring residues connected; while when using cgenff to generate forcefield parameters, I generated parameters for each residue separately without the -H and -OH on two sides. Is that right? I am wondering where is wrong. Thanks a lot.

LZ

I found that the list of atom names and coordinates in the first residue were generated from a different program, thus the atom names are very different from CHARMM. Although using pymol the structure of the first residue is corrct, charmm could not read them correctly. That is the reason for above errors. Under such kind of condition, I am wondering what I can do in order to read the first residue's information correctly. The following is the related charmm input script:

read rtf card name toppar/top_all36_prot.rtf
read param card flex name toppar/par_all36_prot.prm
read rtf card append name toppar/top_all36_cgenff.rtf
read para card flex append name toppar/par_all36_cgenff.prm
stream ./toppar_water_ions.str
stream ./eng_res.str

! Read Sequence of proa
! ----------------------------------------
OPEN READ UNIT 2 CARD NAME lig.pdb
READ SEQU PDB UNIT 2
CLOSE UNIT 2

GENERATE LIG SETUP

OPEN READ UNIT 2 CARD NAME lig.pdb
READ COOR PDB UNIT 2
CLOSE UNIT 2

HBUILD SELEct (hydrogen .AND. .NOT. INITial) END
IC FILL PRESERVE ! define ics from coords where possible
IC PARAMETERS ! fill in any missing bond, angle data
IC BUILD ! place undefined atoms

write coor card name temp.crd

For polymer residues with references to out-of-residue atoms, it is necessary to define terminal patches that add some atoms and fulfill the valence requirements for the heavy atoms involved. See the protein and/or nucleic acid force fields.

In fact, the peptide forms a loop. So, there should be no cterm or n-term.
1).
I used the "PATCHING FIRS NONE LAST NONE" in the str file. I am wondering if you asked me to change this commend. If yes, what I should change it to?

2).Another thing I found is that I got the following warning in the output file:

: No angle parameters for 210 ( C NG32 HGPA)

But I have that one exactly in the str file. So I am wondering why charmm could not recognize it.

Atom names in the coordinate file have to match (be the same as) the names in the residue in the RTF. Use you favorite editor. And, please do consult your supervisor. For a circular peptide you have to use a PATCH to close the circle. There is a PRES to do this for amino acid connections.

In fact, the first 4 residues of the peptides (ABT, ACT, ADT, AET) are engineered residues, and the last two residues (VAL, ALA) are regular residues.
I replaced RESI by PRES for all engineered residues in the str file, and part of the input is like the following:

stream ./toppar_water_ions.str
stream ./eng_res_new.str

read sequence ABT 1
gene new setup
patch link new 1 LIG 1
autogen angl dihe

The following is the error:

CHARMM> read sequence ABT 1

CHARMM> gene new setup

***** ERROR in GENIC ***** Residue 'ABT ' has the wrong patch type

I am wondering how to solve it.

What is the difference between a RESI and a PRES? And how is this related to READ SEQUENCE? The PATCh command is also relevant in this context.