Application of Generative Autoencoder in De Novo Molecular Design

By Thomas Blaschke, Marcus Olivecrona, Ola Engkvist, Jürgen Bajorath, Hongming Chen. 2017

375 citations recorded by the source corpus.

A major challenge in computational chemistry is the generation of novel molecular structures with desirable pharmacological and physiochemical properties. In this work, we investigate the potential use of autoencoder, a deep learning methodology, for de novo molecular design. Various generative autoencoders were used to map molecule structures into a continuous latent space and vice versa and their performance as structure generator was assessed. Our results show that the latent space preserves chemical similarity principle and thus can be used for the generation of analogue structures. Furthermore, the latent space created by autoencoders were searched systematically to generate novel compounds with predicted activity against dopamine receptor type 2 and compounds similar to known active compounds not included in the trainings set were identified.

Subfields: Computational Theory and Mathematics, Materials Chemistry, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Materials Chemistry

In Molecular Biology

Citations in the corpus

Deep Learning in Pharmacogenomics: From Gene Regulation to Patient Stratification (2018, neutral): “design molecules with desired chemical properties [109-111].”
Randomized SMILES strings improve the quality of molecular generative models (2019, background): “Multiple architectures have been reported in literature: the first one is Recurrent Neural Networks (RNNs) [7], but also others such as Variational AutoEncoders (VAEs) [11], Generative Adversarial Networks (GANs) [12, 13], etc.”
The rise of deep learning in drug discovery. (2018, methodology): “[63] utilized VAE to generate novel structures with predicted activity against dopamine receptor type 2.”
The rise of artificial intelligence in healthcare applications (2020, background): “Blaschke et al. also applied adversarial autoenco-ders and Bayesian optimization to generate ligands specific to the dopamine type 2 receptor [22].”
Optimization of Molecules via Deep Reinforcement Learning (2018, background): “uitively, the modication or optimization of a molecule can be done in a step-wise fashion, where each step belongs to one of the following three categories: (1) atom addition, (2) bond addition, and (3) bond removal. The molecule generated is only dependent on the molecule being changed and the modi”
PepCVAE: Semi-Supervised Targeted Design of Antimicrobial Peptide Sequences (2018, background): “[2017], Blaschke et al. [2018], Gómez-Bombarelli et al. [2018], Jin et al. [2018]. Often, the molecule generation task is approached by formulating the design problem as a natural language generation problem, in which molecules are represented as SMILES strings, i.e. sequences of characters. Similar”