DockStream absent from score_component_factory

[j3mdamas]

Hi,

@patronov, I'm tagging you as this is a breaking feature for DockStream.

In the latest release on pip and on the source code on master, the reinvent_scoring does not allow the usage of dockstream as a component on Reinvent 3.0 as the class is not imported and enumerated. When adding such a component to a Reinvent 3.0 run, this is the traceback:

Traceback (most recent call last):
  File "/progs/all/opensource/reinvent/3.0/source/input.py", line 21, in <module>
    manager.run()
  File "/cluster/import/master/progs/all/opensource/reinvent/3.0/source/running_modes/manager.py", line 16, in run
    runner = RunningMode(self.configuration)
  File "/cluster/import/master/progs/all/opensource/reinvent/3.0/source/running_modes/constructors/running_mode.py", line 22, in __new__
    return ReinforcementLearningModeConstructor(configuration)
  File "/cluster/import/master/progs/all/opensource/reinvent/3.0/source/running_modes/constructors/reinforcement_learning_mode_constructor.py", line 20, in __new__
    scoring_function = ScoringFunctionFactory(config.scoring_function)
  File "/progs/all/opensource/reinvent/3.0/miniconda3/lib/python3.7/site-packages/reinvent_scoring/scoring/scoring_function_factory.py", line 16, in __new__
    return cls.create_scoring_function_instance(sf_parameters, scoring_function_registry)
  File "/progs/all/opensource/reinvent/3.0/miniconda3/lib/python3.7/site-packages/reinvent_scoring/scoring/scoring_function_factory.py", line 25, in create_scoring_function_instance
    return scoring_function(parameters, sf_parameters.parallel)
  File "/progs/all/opensource/reinvent/3.0/miniconda3/lib/python3.7/site-packages/reinvent_scoring/scoring/function/custom_product.py", line 14, in __init__
    super().__init__(parameters, parallel)
  File "/progs/all/opensource/reinvent/3.0/miniconda3/lib/python3.7/site-packages/reinvent_scoring/scoring/function/base_scoring_function.py", line 62, in __init__
    self.scoring_components = factory.create_score_components()
  File "/progs/all/opensource/reinvent/3.0/miniconda3/lib/python3.7/site-packages/reinvent_scoring/scoring/score_components/score_component_factory.py", line 69, in create_score_components
    return [self._current_components.get(p.component_type)(p) for p in self._parameters]
  File "/progs/all/opensource/reinvent/3.0/miniconda3/lib/python3.7/site-packages/reinvent_scoring/scoring/score_components/score_component_factory.py", line 69, in <listcomp>
    return [self._current_components.get(p.component_type)(p) for p in self._parameters]
TypeError: 'NoneType' object is not callable

This is quickly fixed with this patch:

@@ -7,7 +7,7 @@ from reinvent_scoring.scoring.score_components import TanimotoSimilarity, \
     SASComponent, MolWeight, PSA, RotatableBonds, HBD_Lipinski, HBA_Lipinski, \
     NumRings, AZlogD74, HLMClint, SlogP, \
     RHClint, HHClint, SolubilityDD, HERG, CACO2Intrinsic, CACO2Efflux, AZdock, RatPKPiP, Top20, GraphLength, \
-    NumberOfStereoCenters
+    NumberOfStereoCenters, DockStream
 from reinvent_scoring.scoring.score_components import BaseScoreComponent
 from reinvent_scoring.scoring.enums import ScoringFunctionComponentNameEnum
 from reinvent_scoring.scoring.score_components.pip.pip_log_prediction_component import PiPLogPredictionComponent
@@ -61,7 +61,8 @@ class ScoreComponentFactory:
             enum.RAT_PK_PIP: RatPKPiP,
             enum.CLAB_TOP_20: Top20,
             enum.RA_SCORE: PiPPredictionComponent,
-            enum.KPUU_PIP: PiPLogPredictionComponent
+            enum.KPUU_PIP: PiPLogPredictionComponent,
+            enum.DOCKSTREAM: DockStream
         }
         return component_map

I can fix it on my deployment, but it's quite cumbersome to start having too many hacks on it. Do you think it is possible to fix it and release a new version of the package?

Thanks in advance, João

[patronov]

Hi João,

Thanks for pointing this out. Will, aim to include this in the next version.

Should also mention that the latest versions (0.0.50) also include changes in ComponentParameters. The fields model_path and smiles are no longer relevant as they are moved to reside within specific_parameters. Didnt update the public git repo but you can get it from pip install reinvent-scoring.

Best,

Atanas

[j3mdamas]

Thanks! So it will be on 0.0.51? Could you let me know so I can update my env file to pick up that version instead when it's available?

[patronov]

Should be there now. Please, also note that the transformation is now in specific_parameters under the keyword: transformation thats where all the transformation related parameters are held.

[j3mdamas]

@patronov thanks a lot! I'll be out of the office, I'll check it when I'm back

[j3mdamas]

I'm currently using 0.0.57 and this is solved. Thanks.