The overall goal of the project is development of approaches to de novo design and structure optimization, their benchmarking and prospective validation. The particular goals are: i) high synthetic accessibility of generated molecules to reduce costs for compound synthesis; ii) high novelty of generated compounds to improve their patentability; iii) high diversity of generated compounds to have a possibility to choose between multiple good suggestions fir further experimental studies; iv) fast generation procedure using a relatively small number of computationally expensive evaluations to save time on the computational stage and speed up the whole discovery pipeline.
Header image
