Molecular design beyond training data with novel extended objective functionals of generative AI models driven by...

Deep generative modeling to stochastically design small molecules is an emerging technology for accelerating drug discovery and development. However, one major issue in molecular generative models is their lower frequency of drug-like compounds. To solve this problem, we develop a novel, quantum annealing, generative-model approach for optimizing deep generative models, an approach that includes integrating with a D-Wave annealing quantum computer. Of particular note, our neural hash function (NHF) is used simultaneously as a regularization scheme and binarization scheme, of which the latter is for transformation between continuous and discrete signals of the classical and quantum neural networks, respectively, in the error evaluation (i.e., objective) function. The compounds generated via the quantum-annealing generative models exhibit higher quality in both validity and drug-likeness than those generated via the fully-classical models, and even exceed the training data in terms of drug-likeness features, without any restraints and conditions to deliberately induce such an optimization. These results suggest that quantum annealing could be used as a stochastic generator integrated with our novel neural network architectures for extending the performance of feature space sampling and the extraction of characteristic features in drug design.