Closed arjun-nufc closed 8 months ago
I should have left more context above. This is a system i had successfully run with 0.10.1, and i used the same yaml file that worked with that version of Perses, from the same directory with the same input files and such.
it looks like self._spectator_md_topologies
is first being defined on line 360 of perses/app/relative_setup.py
, but is being used in self._setup_complex_phase
which is called on line 333.
In contrast, in 0.10.1, self._spectator_md_topologies
is being defined on line 330 and self._setup_complex_phase
is called on line 379. So I think the spectator handling has to be moved before _setup_complex_phase
is called ?
Thank you for this bug report, we will get it fixed before we ship 0.10.3
which will be our last 0.10.x
release (unless we find some nasty bugs) before we switch up our API to something that is easier to maintain.
thanks @mikemhenry !
Thanks for this report. This was introduced by e2ecec6cda982234393dbdca96d38b42f2cc7b7f and it seems this is something that we are not catching in out tests. @arjun-nufc would you be able to provide an example script that is failing so we can have a starting point to make it a test? (It doesn't have to be complete or minimal, just an example we can reproduce so we can make it into a test).
The fix could be just to call the ._setup_complex_phase
method after handling the spectator molecules parameters.
With our 0.10.3 release this issue should be fixed, @arjun-nufc Can you confirm things are working with the latest version? Thanks!
This looks fixed @mikemhenry and @ijpulidos , thank you !
I'm getting the following error with Perses 0.10.2 running a test on a system with a spectator ligand: