The power of deep learning to ligand-based novel drug discovery

ABSTRACT

Introduction

Deep discriminative and generative neural-network models are becoming an integral part of the modern approach to ligand-based novel drug discovery. The variety of different architectures of neural networks, the methods of their training, and the procedures of generating new molecules require expert knowledge to choose the most suitable approach.

Areas covered

Three different approaches to deep learning use in ligand-based drug discovery are considered: virtual screening, neural generative models, and mutation-based structure generation. Several architectures of neural networks for building either discriminative or generative models are considered in this paper, including deep multilayer neural networks, different kinds of convolutional neural networks, recurrent neural networks, and several types of autoencoders. Several kinds of learning frameworks are also considered, including adversarial learning and reinforcement learning. Different types of representations for generating molecules, including SMILES, graphs, and several alternative string representations are also considered.

Expert opinion

Two kinds of problem should be solved in order to make the models built using deep neural networks, especially generative models, a valuable option in ligand-based drug discovery: the issue of interpretability and explainability of deep-learning models and the issue of synthetic accessibility of novel compounds designed by deep-learning algorithms.

KEYWORDS:

• Neural networks
• deep learning
• drug discovery
• generative models
• artificial intelligence

Article Highlights

• The power of deep learning to ligand-based drug discovery stems from the ability to create neural-network filters for virtual screening, to build and run generative neural-network models, and to control the mutation-based generation of chemical structures with desired target properties.

• Deep multilayer neural networks combine and transform initial descriptors to make them more suitable for predicting target properties.

• Deep convolutional neural networks can derive descriptors directly from raw descriptions of chemical structures and use them to predict the target properties of chemical compounds.

• Deep recurrent neural networks can be used to analyze and generate chemical structures represented as SMILES strings.

• Deep autoencoders produce revertible descriptors from which chemical structures can be reconstructed. In conjunction with recurrent neural networks, this allows for designing chemical compounds with desired properties.

• Generative models for performing novel drug discovery can be built using a variety of deep architectures of neural networks in combination with adversarial and reinforcement learning.

This box summarizes key points contained in the article.

Declaration of Interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

I Baskin is funded by the Ministry of Education, Youth and Sports of the Czech Republic (agreement MSMT-5727/2018-2) and the Ministry of Higher Education and Science of the Russian Federation (project number: RFMEFI58718X0049).