choderalab/mhc

Examples for MHC-peptide free energy calculations using YANK

mhc

Examples for MHC-peptide free energy calculations using YANK

mhc's Issues

Does this work with the current version of Yank? Ligand resname error.

I'm able to run other Yank examples (t4-lysozyme), but for this example I get an error that has something to do with the ligand resname. It appears to be trying to open the ligand as a mol2 file, but it's a peptide in PDB format. I am also running into the same error when I try to modify the t4-lysozyme example to use a peptide ligand.

2020-01-28 09:44:23,673: Setting CUDA platform to use precision model 'mixed'.
2020-01-28 09:44:23,772: WARNING - openmmtools.multistate.multistatesampler - Warning: The openmmtools.multistate API is experimental and may change in future releases
2020-01-28 09:44:23,813: Setting up the systems for mhc and peptide using solvent pme
Traceback (most recent call last):
File "run_yank.py", line 150, in
main(args.yaml_script_path, args.job_id, args.n_jobs)
File "run_yank.py", line 138, in main
experiment_builder.run_experiments()
File "/home/comer/anaconda3/envs/yank/lib/python3.7/site-packages/yank/experiment.py", line 751, in run_experiments
completed[exp_index] = self._run_experiment(exp, write_status=write_status)
File "/home/comer/anaconda3/envs/yank/lib/python3.7/site-packages/yank/experiment.py", line 3247, in _run_experiment
built_experiment = self._build_experiment(experiment_path, experiment)
File "run_yank.py", line 126, in _build_experiment
experiment = super(TwoRestraintsBuilder, self)._build_experiment(*args, **kwargs)
File "/home/comer/anaconda3/envs/yank/lib/python3.7/site-packages/yank/experiment.py", line 3042, in _build_experiment
ligand_resname = utils.Mol2File(ligand_descr['filepath']).resname
File "/home/comer/anaconda3/envs/yank/lib/python3.7/site-packages/yank/utils.py", line 1124, in resname
return next(self.resnames)
StopIteration

More general ligand atom identification

Current DSL strings in the .yaml files to identify ligand in both complex and solvation experiments are highly specific. A more general way is needed to demarcate ligands to increase automated throughput.

keepIds

Attempting to run PDBFixer with keepIds=True leads to instability during subsequent short OpenMM simulation, with error message "Particle coordinates are NaN."