Realistic molecule optimization on a learned graph manifold
This work addresses the issue of unrealistic molecule generation in de novo drug design, which is an incremental improvement over existing methods.
The paper tackled the problem of generating unrealistic molecules in deep learning-based molecular graph optimization by introducing a hybrid approach combining an autoregressive model with Metropolis algorithm sampling, resulting in a method called learned realism sampling (LRS) that produces more realistic molecules and outperforms recent baselines in optimization with similarity constraints.
Deep learning based molecular graph generation and optimization has recently been attracting attention due to its great potential for de novo drug design. On the one hand, recent models are able to efficiently learn a given graph distribution, and many approaches have proven very effective to produce a molecule that maximizes a given score. On the other hand, it was shown by previous studies that generated optimized molecules are often unrealistic, even with the inclusion of mechanics to enforce similarity to a dataset of real drug molecules. In this work we use a hybrid approach, where the dataset distribution is learned using an autoregressive model while the score optimization is done using the Metropolis algorithm, biased toward the learned distribution. We show that the resulting method, that we call learned realism sampling (LRS), produces empirically more realistic molecules and outperforms all recent baselines in the task of molecule optimization with similarity constraints.