Open aishiba0721 opened 10 months ago
Sorry for the late of response! Regarding a given molecule, assuming its SMILES is CC1=CC=C(C=C1)C2=CC(=NN2C3=CC=C(C=C3)S(=O)(=O)N), which is also the SMILES for Celecoxib. You can create the MolFragTree object defined in TreeInvent. The MolFragTree.get_CL_states() attribute function allows direct access to the topological structures at each step during the generation process. The Gclique_To_Constrain() function can be used to obtain the topological constraints corresponding to the molecule, facilitating scaffold hopping and reinforcement learning based on a reference molecule. The example script (also a reinforcement learning script for topological constraints on Celecoxib) and the configuration file are attached. Celecoxib.zip
Hello, I am currently validating some data from the paper using this project. I have encountered some issues during the reproduction of the case depicted in the attached image. Could you please provide a reference JSON configuration file to guide me on how to correctly define the skeleton in the graph? Your assistance would be greatly appreciated.