Error while Dimer:Dimer mmgbsa calculation

[Krish]

Hi All

In my system there are four chains and each represent a monomer. The protein X1 is a dimer which binds to X2 which also a dimer. Chain A (one monomer of protein X1) binds to Chain C (one monomer of protein X2). Then Chain B (another monomer of protein X1) binds to Chain D (another monomer of protein X2). I am trying to calculate MM/GBSA for a dimer complex with another dimer. I am getting below error. I have no problem in calculation when I am considering each monomer binding energy. 

[INFO   ] Starting gmx_MMPBSA v1.6.0

[INFO   ] Command-line

  gmx_MMPBSA -O -i mmpbsa.in -cs 80ns.pdb -ci index.ndx -cg 10 11 -ct fit-mmgbsa.pdb -o FINAL_RESULTS_MMPBSA.dat -eo FINAL_RESULTS_MMPBSA.csv -do FINAL_DECOMP_MMPBSA.dat -deo FINAL_DECOMP_MMPBSA.csv

[INFO   ] Checking mmpbsa.in input file...

[INFO   ] Checking mmpbsa.in input file...Done.

[INFO   ] Checking external programs...

[INFO   ] cpptraj found! Using /software/ambertools-21/bin/cpptraj

[INFO   ] tleap found! Using /software/ambertools-21/bin/tleap

[INFO   ] parmchk2 found! Using /software/ambertools-21/bin/parmchk2

[INFO   ] sander found! Using /software/ambertools-21/bin/sander

[INFO   ] elsize found! Using /software/ambertools-21/bin/elsize

[INFO   ] Using GROMACS version > 5.x.x!

[INFO   ] gmx found! Using /software/gromacs-2021-cuda11/bin/gmx

[INFO   ] Checking external programs...Done.

[INFO   ] Building AMBER topologies from GROMACS files...

[INFO   ] Get PDB files from GROMACS structures files...

[INFO   ] Making gmx_MMPBSA index for complex...

[INFO   ] Normal Complex: Saving group a_1-7780_a_7781-13142 (10_11) in _GMXMMPBSA_COM_index.ndx file as _GMXMMPBSA_COM.pdb

[INFO   ] No receptor structure file was defined. Using ST approach...

[INFO   ] Using receptor structure from complex to generate AMBER topology

[INFO   ] Normal Receptor: Saving group a_1-7780 (10) in _GMXMMPBSA_COM_index.ndx file as _GMXMMPBSA_REC.pdb

[INFO   ] No ligand structure file was defined. Using ST approach...

[INFO   ] Using ligand structure from complex to generate AMBER topology

[INFO   ] Normal Ligand: Saving group a_7781-13142 (11) in _GMXMMPBSA_COM_index.ndx file as _GMXMMPBSA_LIG.pdb

[INFO   ] Checking the structures consistency...

[INFO   ] 

[INFO   ] Generating AMBER Compatible PDB Files...

[INFO   ] Changing the Complex residues name format from GROMACS to AMBER...

[INFO   ] Changing the Receptor residues name format from GROMACS to AMBER...

[INFO   ] Changing the Ligand residues name format from GROMACS to AMBER...

[INFO   ] Splitting  receptor and ligand in PDB files..

[INFO   ] Building tleap input files...

[INFO   ] Selecting residues by distance (4 Å) between receptor and ligand for decomposition analysis...

[INFO   ] Selected 72 residues:

R:A:LYS:52  R:A:ASN:53  R:A:ARG:59  R:A:ARG:63  R:A:LYS:123 R:A:ARG:130 R:A:TYR:131 R:A:GLY:172 R:A:LEU:175 R:A:ASN:176

R:A:VAL:179 R:A:TYR:182 R:A:GLU:183 R:A:LEU:206 R:A:ASP:207 R:A:LEU:209 R:A:GLU:211 R:A:TYR:214 R:A:LYS:215 R:A:LEU:219

R:A:ILE:220 R:A:LEU:223 R:A:LEU:230 R:B:LYS:52  R:B:ARG:59  R:B:ARG:63  R:B:ARG:130 R:B:TYR:131 R:B:LEU:175 R:B:ASN:176

R:B:VAL:179 R:B:TYR:182 R:B:LYS:215 R:B:THR:218 R:B:LEU:219 R:B:GLN:222 R:B:LEU:223 R:B:ASP:226 R:B:ASN:227 R:B:LEU:230

R:B:TRP:231 R:B:ASP:234 L:C:PHE:17  L:C:ARG:18  L:C:ARG:19  L:C:ALA:20  L:C:VAL:21  L:C:SEP:22  L:C:LEU:24  L:C:GLN:28

L:C:PHE:37  L:C:ARG:41  L:C:LEU:44  L:C:ILE:45  L:C:ALA:48  L:C:ARG:49  L:C:ARG:52  L:D:LYS:15  L:D:GLY:16  L:D:ARG:18

L:D:ARG:19  L:D:ALA:20  L:D:VAL:21  L:D:SEP:22  L:D:GLU:23  L:D:LEU:24  L:D:ALA:29  L:D:ILE:32  L:D:MET:33  L:D:PHE:37

L:D:ILE:38  L:D:ARG:41

[INFO   ] Cleaning normal complex trajectories...

[INFO   ] Building AMBER topologies from GROMACS files... Done.

[INFO   ] Loading and checking parameter files for compatibility...

  File "/software/gmx_mmpbsa-1.6/bin/gmx_MMPBSA", line 33, in <module>

    sys.exit(load_entry_point('gmx-MMPBSA==1.6.0', 'console_scripts', 'gmx_MMPBSA')())

  File "/software/gmx_mmpbsa-1.6/lib/python3.9/site-packages/./gmx_MMPBSA-1.6.0-py3.9.egg/GMXMMPBSA/app.py", line 100, in gmxmmpbsa

    app.file_setup()

  File "/software/gmx_mmpbsa-1.6/lib/python3.9/site-packages/./gmx_MMPBSA-1.6.0-py3.9.egg/GMXMMPBSA/main.py", line 127, in file_setup

    create_inputs(INPUT, self.normal_system, self.pre)

  File "/software/gmx_mmpbsa-1.6/lib/python3.9/site-packages/./gmx_MMPBSA-1.6.0-py3.9.egg/GMXMMPBSA/createinput.py", line 109, in create_inputs

    com_mdin = SanderGBDecomp(com_input, rec_res, lig_res, pri_res)

  File "/software/gmx_mmpbsa-1.6/lib/python3.9/site-packages/./gmx_MMPBSA-1.6.0-py3.9.egg/GMXMMPBSA/createinput.py", line 841, in __init__

    self.make_mdin()

  File "/software/gmx_mmpbsa-1.6/lib/python3.9/site-packages/./gmx_MMPBSA-1.6.0-py3.9.egg/GMXMMPBSA/createinput.py", line 322, in make_mdin

    self.mdin.change(self.parent_namelist[key], key, self.INPUT[self.namelist][self.name_map[key]])

  File "/software/ambertools-21/lib/python3.9/site-packages/parmed/amber/mdin/mdin.py", line 293, in change

    self.cntrl_nml[variable] = mytype(value)

ValueError: could not convert string to float: 'failed.'

Exiting. All files have been retained.

I have also tried entire protein X1 a single chain and X2 single chain, then I have also another error i.e. in building tleap is unsuccessful. Because tleap is considering the bond between end of the each monomer of X1 and X2.

Thanks in advance!

Kind regards

Krishna

[marioe911116]

Please, install ParmEd from our repository and try again...

python -m pip install git+https://github.com/Valdes-Tresanco-MS/ParmEd.git@v3.4 -U

[Krish]

The parmed: Version 3.4.1 already installed.

Thanks

Krishna

[marioe911116]

You should have ParmEd 3.4.3+9.gb81644c installed from our repo

[Krish]

I think we have installed it. Still I will ask my computer admin. For rest of the calculations there is no error. For the same system monomer vs monomer is working fine. I checked the gmx_MMGBSA_test is also fine. I calculated various protein-lig systems and other 3 protein-protein interactions working fine. 

[Mario Sergio Valdes]

The recommended version of ParmEd is 3.4.3+9.gb81644c because we fixed several problems in our fork.

To install this version, you must uninstall the current version of ParmEd (python -m pip uninstall parmed) and then install the version of our fork (python -m pip install git+https://github.com/Valdes-Tresanco-MS/ParmEd.git@v3.4 -U)

You can check the ParmEd version in the first section of the gmx_MMPBSA.log file.

With respect to the tleap error. Please, define the -cp topol.top option, the topology generation from a structure is deprecated

[marioe911116]

if thats the case, can you share the files you are using?

[marioe911116]

As per discussion offline:

The pdbs look broken probably because of PBC conditions or something like that. Nevertheless, the program works fine when alpb=0 (deactivate alpb option). The thing is that when alpb is on, the "electrostatic radius" is calculated from a generated pqr file. Seems the program we used to do so doesn't like this broken pdb and reported an error. 

[Krish]

It was not problem of broken pdb it was rather "alpb". My system didn't like the "alpb" when there was multiple phosphoserines (charge -2e), but without phosphoserines it worked fine. Anyway problem solved. Thanks Mario for help.