ccpbiosim / fesetup Goto Github PK

Good day! I hope you're well.

Everything worked for me, up until the perturbations when I run morph.in. PLEASE HELP!
I am getting the following, in the terminal and leap.log file:

Terminal:
Making biomolecule 181L...
Morphs will be generated for Sire
Morphing benzene to o-xylene...
ERROR: benzeneo-xylene failed: Leap did not create the topology and/or coordinate file(s): vacuum.parm7, vacuum.rst7
Morphing o-xylene to benzene...
Creating complex 181L:benzene with ligand morph benzeneo-xylene...
AttributeError: 'NoneType' object has no attribute 'edit'

leap.log:

s = combine {cmp0}
Sequence: LIG
saveAmberParm s "state0.parm7" "state0.rst7"
Checking Unit.
Building topology.
Building atom parameters.
For atom: .R<LIG 1>.A<DU13 13> Could not find vdW (or other) parameters for type: du
For atom: .R<LIG 1>.A<DU14 14> Could not find vdW (or other) parameters for type: du
For atom: .R<LIG 1>.A<DU15 15> Could not find vdW (or other) parameters for type: du
For atom: .R<LIG 1>.A<DU16 16> Could not find vdW (or other) parameters for type: du
For atom: .R<LIG 1>.A<DU17 17> Could not find vdW (or other) parameters for type: du
For atom: .R<LIG 1>.A<DU18 18> Could not find vdW (or other) parameters for type: du
Parameter file was not saved.
savepdb s "state0.pdb"
Writing pdb file: state0.pdb
quit
Quit

The mass assigned to dummy atoms needs fixing. The Lennard-Jones terms need to be checked.

The options lig['correct_for_pH'] and lig['pH'] are not documented.

1. check how/if they work
2. fix any bugs
3. add documentation

FESetup1.2 uses ambertools16. Update code so it works with newer versions of ambertools.

Unit tests need to be written.

When trying to compute the morph between two ligands FESetup struggles to find the maximum common substructure and will therefore propose a buggy morph.

The first image shows the two pdb files of the ligands to be morphed.
The next shows the resulting morphed ligand:

The identified MCS looks like this:

FESetup input looks like this:

FE_type = Sire
AFE.separate_vdw_elec = False
mcs.match_by = spatially-closest
mcs.timeout = 0

[globals]
forcefield = amber, ff14SB, tip3p
gaff = gaff2
logfile = CatS_30.log

[ligand]
basedir = poses
file.name = ligand.pdb
molecules = CatS_30, CatS_29

box.type = rectangular
box.length = 12.0
neutralize = yes

min.nsteps = 1000
min.restr_force = 10.0
min.restraint = notsolvent

md.heat.nsteps = 1000
md.heat.restr_force = 10.0
md.heat.restraint = notsolvent

md.constT.nsteps = 1000
md.constT.restr_force = 10.0
md.constT.restraint = notsolvent

md.press.T = 298.0
md.press.nsteps = 20000
md.press.p = 1.0

morph_pairs = CatS_29 > CatS_30

[protein]
basedir = protein
file.name = protein.pdb
molecules = cats

[complex]
pairs = cats : CatS_30

box.type = rectangular

The zip archive contains all relevant coordinate files.

FESetup.zip

In prepare/amber there are three files (complex.py, ligand.py, and protein.py) which all contain prepare_top and create_top functions. These functions should be merged to avoid duplication and reduce the confusion.

As reported by Laura Perez. When a morph involves both insertion and deletion of atoms, for GROMACS this needs to be done in two steps. FESetup is creating morph1 and morph2 files, but creating charges on some dummy atoms so that the removal and insertion of atoms is being mixed rather than as two separate steps. This is leading to unstable simulations.

FESetup struggles to get the correct charge for positive and negatively charged ligands. It seems like this can also cause trouble with the complexes.

Upgrade code from python2.7 to python3.7

Hi there,
I would like to setup a simulation with a ligand and a cofactor. Is it possible with FESetup?

Bests

Cesar

Dear Developers of FES-setup,

I just read about your program, and I am excited to try it out. I was wondering if it works for residue mutations in proteins (with no ligand), i.e. to find the difference in free energy of a solvated protein system when for example a glycine residue has been mutated to a serine?

Thanks in advance,
Holly Freedman

Hi,
It's a very useful tool! While the internal ambertools16 is somewhat old.
How can I use it with newer amertools?
I installed the tool using executable file from http://www.ccpbiosim.ac.uk/software.
I also noticed this [https://github.com//issues/5], and tried to modify the installed package.
While it doesn't work, and errior dettails are as follows:

`Making biomolecule rec...
ERROR: rec failed: Leap did not create the topology and/or coordinate file(s): vacuum.parm7, vacuum.rst7
Making ligand lig1...
IOError: [Errno 2] No such file or directory: 'sqm.out'
`

Could someone give me a brief guidance?

The output produced by FESetup should be reorganised so that it flows more naturally into running simulations.

Hi, I have noticed that when running FESetup, in the part of complexation it brings the protein.pdb to the folder instead of the protonated.pdb. If this is true then pH is not correct. What is pH by default? Is there any way to take the protonated option?

Hi, I am running a Mac and have been trying to install FESetup from the downloadable shell script from https://www.ccpbiosim.ac.uk/software

When running the script it errors due to my Mac OS:

`Installing package FESetup release 1.2.1...

Determining system environment...
ERROR: unsupported operating system: Darwin.
Aborting the installation of FESetup release 1.2.1!`

I am guessing the shell script works only for Linux and not MacOS. Are there any plans to release a Mac compatible version anytime soon?

Thanks

What happened to the wiki documentation since the migration?

I am not sure if this is intended, but in order to overcome poor mapping one way of doing so is giving an explicit mapping. if I use atom indexes from the input pdb file the explicit mapping will still fail, because FESetup changed the atom indexing in the parametrisation. So only once I have generated a set of vacuum.rst7 and vacuum.parm7 files is it possible to do explicit mapping with the atom indexes in these files.

Input file before parametrisation with atom indexes:

Atom indexes after parametrisation:

Hello,

I would like to run a simulation of a ligand binding to a protein with a co-factor. I see in an earlier post (December 14-16) it was stated that one could run with a co-factor by including it with the protein. Could you be more specific on exactly how one includes the co-factor in the set-up process?

Right now, I am getting an error from tleap, which I have copied below. I think this is because it is looking to assign protein parameters to the small molecule co-factor? Do I have to pre-parametrize the co-factor and then add it in as a user parameter file? Do I have to split the pdb file into one of the protein and one of the co-factor and then combine them later? Thanks in advance for your help. I can attach the .in file or other files I am using upon request.

FATAL: Atom .R<3M3 461>.A<FAC 18> does not have a type.
FATAL: Atom .R<3M3 461>.A<OAM 17> does not have a type.
FATAL: Atom .R<3M3 461>.A<NAL 16> does not have a type.
FATAL: Atom .R<3M3 461>.A<NAK 15> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAR 14> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAQ 13> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAP 12> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAO 11> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAN 10> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAJ 9> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAI 8> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAH 7> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAG 6> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAF 5> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAE 4> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAD 3> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAB 2> does not have a type.
FATAL: Atom .R<3M3 461>.A<CAA 1> does not have a type.