Molecular Generative Model Based On Adversarially Regularized Autoencoder
This work addresses the challenge of generating novel and valid molecules for drug discovery, representing an incremental improvement by combining elements of VAE and GAN.
The authors tackled the problem of molecular design by proposing a new generative model, the adversarially regularized autoencoder (ARAE), which outperformed conventional models in validity, uniqueness, and novelty per generated molecule, and demonstrated successful conditional generation of drug-like molecules, including EGFR inhibitors.
Deep generative models are attracting great attention as a new promising approach for molecular design. All models reported so far are based on either variational autoencoder (VAE) or generative adversarial network (GAN). Here we propose a new type model based on an adversarially regularized autoencoder (ARAE). It basically uses latent variables like VAE, but the distribution of the latent variables is obtained by adversarial training like in GAN. The latter is intended to avoid both inappropriate approximation of posterior distribution in VAE and difficulty in handling discrete variables in GAN. Our benchmark study showed that ARAE indeed outperformed conventional models in terms of validity, uniqueness, and novelty per generated molecule. We also demonstrated successful conditional generation of drug-like molecules with ARAE for both cases of single and multiple properties control. As a potential real-world application, we could generate EGFR inhibitors sharing the scaffolds of known active molecules while satisfying drug-like conditions simultaneously.