Generative Network Complex for the Automated Generation of Drug-like Molecules

Abstract

Current drug discovery is expensive and time-consuming. It remains a challenging task to create a wide variety of novel compounds that not only have desirable pharmacological properties but also are cheaply available to low-income people. In this work, we develop a generative network complex (GNC) to generate new drug-like molecules based on the multiproperty optimization via the gradient descent in the latent space of an autoencoder. In our GNC, both multiple chemical properties and similarity scores are optimized to generate drug-like molecules with desired chemical properties. To further validate the reliability of the predictions, these molecules are reevaluated and screened by independent 2D fingerprint-based predictors to come up with a few hundreds of new drug candidates. As a demonstration, we apply our GNC to generate a large number of new BACE1 inhibitors, as well as thousands of novel alternative drug candidates for eight existing market drugs, including Ceritinib, Ribociclib, Acalabrutinib, Idelalisib, Dabrafenib, Macimorelin, Enzalutamide, and Panobinostat.

Merging Ligand-Based and Structure-Based Methods in Drug Discovery: An Overview of Combined Virtual Screening Approaches

“…The design of target chemical libraries is actually an undertaking of increasing interest, as illustrated by a number of recent studies that have reported the implementation of artificial intelligence-based algorithms [ 63 , 64 , 65 , 183 , 184 , 185 , 186 ]. One of them is the development of ReLeaSE (Reinforcement Learning for Structural Evolution), which integrates a generative deep neural network with a predictive one into a joint framework for the design of novel compounds satisfying certain chemical requirements, as illustrated with the biased selection of compounds fulfilling a specific range of physical properties (i.e., melting temperature and lipophilicity) or inhibitory activity against the desired target protein (Janus protein kinase 2) [ 62 ].…”

Advances in De Novo Drug Design: From Conventional to Machine Learning Methods

“…The use of CADD algorithms and tools could reduce drug development costs and time significantly with conservative estimates suggesting AI pipelines require less than 1/3 of the current time and cost [152,153]. Examples of DRL-based de novo drug design include the development of adenosine A2A receptor ligands [83], rapid identification of potent DDR1 kinase inhibitors [154], and the development of a large number of new BACE1 inhibitors, which is an enzyme involved in Alzheimer's disease [118].…”

“…Thus, the latent vector is an intermediate representation containing the "meaning" of the input sequence. In the case of de novo drug design, the input and output sequences are both SMILES strings [118]. A generative network complex was successful in generating new drug-like molecules based on multi-property optimization via a gradient descent in the latent space of an autoencoder [118].…”

“…In the case of de novo drug design, the input and output sequences are both SMILES strings [118]. A generative network complex was successful in generating new drug-like molecules based on multi-property optimization via a gradient descent in the latent space of an autoencoder [118]. In this approach, both multiple chemical properties and similarity scores were optimized to generate drug-like molecules with the desired properties.…”

“…The predictions of this method were validated using independent two-dimensional predictors based on molecular fingerprints. Finally, the method was utilized to generate a large number of new BACE1 inhibitors, as well as thousands of novel alternative drug candidates for eight existing drugs currently on the market, including Ceritinib, Ribociclib, Acalabrutinib, Idelalisib, Dabrafenib, Macimorelin, Enzalutamide, and Panobinostat [118]. A seq2seq AE was also used to develop a de novo drug-design approach using SMILES strings [119].…”