[BUG] Failure to run when minimising with restraints using OpenMM

[mb2055]

Describe the bug
Same issue as #339 , this time in OpenMM when using Protocol.Minimisation and Protocol.Equilibration with restraints.

To Reproduce
The following code fails to produce a system, with process_m1.getSystem() returning None

import BioSimSpace as BSS

ethane = BSS.Parameters.gaff("CC").getMolecule()

methanol = BSS.Parameters.gaff("CO").getMolecule()

mapping = BSS.Align.matchAtoms(ethane, methanol)

ethane = BSS.Align.rmsdAlign(ethane, methanol, mapping)

merged = BSS.Align.merge(ethane, methanol, mapping)

vac = merged.toSystem()

minimise = BSS.Protocol.Minimisation(restraint="heavy")

process_m1 = BSS.Process.OpenMM(vac, minimise, work_dir="Minimise_Vacuum")

process_m1.start()
process_m1.wait()

removing restraint="heavy" solves the issue

(please complete the following information):

• OS :Ubuntu 22.04.5 LTS
• Version of Python: 3.12.4
• Version of BioSimSpace: 2024.3.0.dev+18.gdc6602e4
• I confirm that I have checked this bug still exists in the latest released version of BioSimSpace: yes

[mb2055]

Running openmm_script.py directly highlights the folowing line as the issue:
ref_prm = parmed.load_file('openmm.prm7', 'openmm_ref.rst7')
with the error raised being cannot reshape array of size 0 into shape (0,3).

Attempting to load openmm.prm7 and openmm_ref.rst7 with BioSimSpace gives the following:

In [1]: import BioSimSpace as BSS

INFO:numexpr.utils:Note: NumExpr detected 20 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8.
INFO:numexpr.utils:NumExpr defaulting to 8 threads.

In [2]: BSS.IO.readMolecules(["openmm.prm7", "openmm_ref.rst7"])
╭───────────────────── Traceback (most recent call last) ──────────────────────╮
│ in <module>:1                                                                │
│                                                                              │
│ /home/matt/code/biosimspace/python/BioSimSpace/IO/_io.py:550 in              │
│ readMolecules                                                                │
│                                                                              │
│    547 │   │   │   │   if _isVerbose():                                      │
│    548 │   │   │   │   │   raise IOError(msg) from e0                        │
│    549 │   │   │   │   else:                                                 │
│ ❱  550 │   │   │   │   │   raise IOError(msg) from None                      │
│    551 │   │   │   else:                                                     │
│    552 │   │   │   │   msg = "Failed to read molecules from: %s" % files     │
│    553 │   │   │   │   if _isVerbose():                                      │
╰──────────────────────────────────────────────────────────────────────────────╯
OSError: Incompatibility between molecular information in files: ['openmm.prm7',
'openmm_ref.rst7']

[lohedges]

That's strange. Presumably the LJ change is causing a self-consistency issue somewhere. I'll try to debug this tomorrow.

[lohedges]

I see the issue. If None is passed as the reference system then is just takes the system as a reference. However, it doesn't have any logic to deal with perturbable systems, so doesn't extract the end state. As such, the regular coordinates property is used and no atoms are found. Will fix this now.