Molecular generative model based on conditional variational autoencoder for de novo molecular design

Abstract

We propose a molecular generative model based on the conditional variational autoencoder for de novo molecular design. It is specialized to control multiple molecular properties simultaneously by imposing them on a latent space. As a proof of concept, we demonstrate that it can be used to generate drug-like molecules with five target properties. We were also able to adjust a single property without changing the others and to manipulate it beyond the range of the dataset.Electronic supplementary materialThe online version of this article (10.1186/s13321-018-0286-7) contains supplementary material, which is available to authorized users.

Deep learning for molecular design—a review of the state of the art

“…In Type 2 VAEs, property data y is embedded directly into the latent space during training. 87,110 Supervised and semi-supervised VAEs can both be used for conditional sampling, and thus are sometimes called "conditional VAEs". In the traditional way of doing conditional sampling, y is specified and then one samples from the prior p(z).…”

Constrained Bayesian optimization for automatic chemical design using variational autoencoders

“…The literature concerning generative models of molecules has exploded since the first work on the topic Gómez-Bombarelli et al [2018]. Current methods feature molecular representations such as SMILES [Janz et al, 2018, Segler et al, 2017, Skalic et al, 2019, Ertl et al, 2017, Lim et al, 2018, Kang and Cho, 2018, Sattarov et al, 2019, Gupta et al, 2018, Harel and Radinsky, 2018, Yoshikawa et al, 2018, Bjerrum and Sattarov, 2018, Mohammadi et al, 2019 and graphs [Simonovsky and Komodakis, 2018, Li et al, 2018a, De Cao and Kipf, 2018, Kusner et al, 2017, Dai et al, 2018, Samanta et al, 2019, Li et al, 2018b, Kajino, 2019, Jin et al, 2019, Bresson and Laurent, 2019, Lim et al, 2019, Pölsterl and Wachinger, 2019, Krenn et al, 2019, Maziarka et al, 2019, Madhawa et al, 2019, Shen, 2018, Korovina et al, 2019 In this section we conduct an empirical test of the hypothesis from [Gómez-Bombarelli et al, 2018] that the decoder's lack of efficiency is due to data point collection in "dead regions" of the latent space far from the data on which the VAE was trained. We use this information to construct a binary classification Bayesian Neural Network (BNN) to serve as a constraint function that outputs the probability of a latent point being valid, the details of which will be discussed in the section on labelling criteria.…”

The Trends of De Novo Molecular Designs in the Twenty-First Century: A Mini-Review

“…Molecules in the library may not meet the given criteria. In this case, traditional optimization methods such as a genetic algorithm can be used to enhance further molecular properties beyond the requirements by structural modifications (Cheng, Li, Zhou, Wang, & Bryant, Published by SCHOLINK INC. 2012; Reymond, van Deursen, Blum, & Ruddigkeit, 2010;Lim et al, 2018). However, they have a fundamental limitation in terms of efficiency because many trials and errors are inevitable to optimize molecular properties in a substantial molecular space.…”

“…The approach skips the high throughput virtual screening step and uses deep learning-based generative models directly to fabricate molecules having specific target properties. A condition vector is incorporated into the system, which regulates the target properties simultaneously when exposed to a particular environment (Lim et al, 2018). The group demonstrated that it was possible to induce five target properties (MW, LogP, HBD, HBA, and TPSA) having an error range of 10%.…”