Hello all,
I've been attempting to use the MCPB.py tool-chain to build amber
parameters for a iron porphyrin IX, heme, system, and have been having a
few difficulties that may be related.
Like others in the past, I optimized a truncated system lacking the
propionate groups and bound to imidazole and carbon monoxide ligand. This
was optimized with using restricted DFT 631G** B3LYP, which worked just
fine.
>From there I added the propionate groups and produced the MK charges from
an energy calculation at the same theory level.
This was all good so I started with MCPB.py since Amber15 had just come out
and it appeared to be the best option for generating accurate parameters
for the iron center.
I made the PDB file for the whole system from the optimized structure and
the iron / naa mol2 files from those produced with the MK charges using
UCSF Chimera maintaining the input atom types. I had some error with the
mol2 files due to formatting differences but I ran them through antechamber
to produce files that seemed to work correctly. I used the mol2 files and
antechamber to also generate the starting frcmod file.
Then I ran MCPB.py with my input file (g09 was specified) and there was a
key error called for 'FE', I looked through and tried altering the
capitalization which resulted in a error called for 'HEM-Fe' which then
somehow decided afterwards to be simply 'Fe' in the next attempt.
Gaussian input files were still produced so I thought to run it anyway. the
iron had been changed to a flourine so I corrected this and let it run
overnight on the cluster while I dozed. I had modified the optimization to
an energy calculation at the previous level of theory because the problems
the propionates would produce otherwise (pesky charged groups) and read the
esp fitting checkpoint file in as an initial guess to speed things along.
The freq input was also fixed to use the same level of theory as everything
else.
This was successful.. but then when I ran step 2 using both z and s
options... it kicked back another key error when writing the pre.frcmod
file for 'F-2'.
I am not sure how to proceed, but there is definitely a problem with some
formatting involving my iron core.
Output from step 1:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
The following is the input variable you have:
The variable original_pdb is : imz-carboxyheme-1.pdb
The variable ion_ids is : [9]
The variable ion_mol2files is : ['imz-carboxyheme-1_fe.mol2']
The variable group_name is : hem
The variable cut_off is : 2.8
The variable ionchg_fixation is : 0
The variable gau_version is : g09
The variable sqm_opt is : 0
The variable force_field is : ff14SB
The variable gaff is : 1
The variable frcmodfs is : ['imz-carboxyheme-1_ph.frcmod']
The variable naa_mol2files is : ['imz-carboxyheme-1_ph.mol2']
The variable water_model is : TIP3P
The variable ion_paraset is : CM (Only for nonbonded model)
The variable ion_info is : []
******************************************************************
* *
*=======================Metal Site Information===================*
* *
******************************************************************
***Selected Metal ion Fe is atom 9 in residue 1-HEM
1-HEM.N2 is in 2.8 Angstrom of 1-HEM.N2
1-HEM.N3 is in 2.8 Angstrom of 1-HEM.N3
1-HEM.N4 is in 2.8 Angstrom of 1-HEM.N4
1-HEM.N5 is in 2.8 Angstrom of 1-HEM.N5
1-HEM.N6 is in 2.8 Angstrom of 1-HEM.N6
***The following residues are in the Metal Site:
Residue 1-HEM
***The large model contains the following residues:
[1]
******************************************************************
* *
*=======================Building models==========================*
* *
******************************************************************
***Creating the sidechain model...
It contains the residue 1-HEM
Totally there are 88 atoms in the sidechain model.
Totally there are 359 electrons in the sidechain model.
***Creating the standard model...
It contains the residue 1-HEM
Totally there are 88 atoms in the standard model.
***Creating the large model...
Creating the normal residue 1-HEM
Totally there are 88 atoms in the large model.
Totally there are 359 electrons in the large model.
Traceback (most recent call last):
File "/yggdrasil/amber14/bin/MCPB.py", line 419, in <module>
watermodel, 2, sqmopt)
File
"/yggdrasil/amber14/lib/python2.7/site-packages/mcpb/gene_model_files.py",
line 1167, in gene_model_files
ionids, chargedict, totchg, outf, watermodel, sqmopt)
File
"/yggdrasil/amber14/lib/python2.7/site-packages/mcpb/gene_model_files.py",
line 984, in build_large_model
chg = str(int(chargedict[i]))
KeyError: 'Fe'
Output from step 2:
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
The following is the input variable you have:
The variable original_pdb is : imz-carboxyheme-1.pdb
The variable ion_ids is : [9]
The variable ion_mol2files is : ['imz-carboxyheme-1_fe.mol2']
The variable group_name is : hem
The variable cut_off is : 2.8
The variable ionchg_fixation is : 0
The variable gau_version is : g09
The variable sqm_opt is : 0
The variable force_field is : ff14SB
The variable gaff is : 1
The variable frcmodfs is : ['imz-carboxyheme-1_ph.frcmod']
The variable naa_mol2files is : ['imz-carboxyheme-1_ph.mol2']
The variable water_model is : TIP3P
The variable ion_paraset is : CM (Only for nonbonded model)
The variable ion_info is : []
******************************************************************
* *
*===================Generate the Initial frcmod file=============*
* *
******************************************************************
Atoms which has changed the atom types: [5, 9, 14, 22, 30, 38]
1-HEM.5-N2 : n2 --> X1
1-HEM.9-Fe : Fe --> Z1
1-HEM.14-N3 : n2 --> X2
1-HEM.22-N4 : n2 --> X3
1-HEM.30-N5 : n2 --> X4
1-HEM.38-N6 : n2 --> X5
Traceback (most recent call last):
File "/yggdrasil/amber14/bin/MCPB.py", line 446, in <module>
gaff, frcmodfs, watermodel)
File
"/yggdrasil/amber14/lib/python2.7/site-packages/mcpb/gene_pre_frcmod_file.py",
line 369, in gene_pre_frcmod_file
rmin = IonLJParaDict[element + chg][0]
KeyError: 'F-2'
Thanks everybody!
-David Poole
M.Sc. San Francisco State University
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 19 2015 - 11:00:03 PDT
