Discoidin domain receptor 1 (DDR1) inhibitors with a desired pharmacophore were designed using deep generative models (DGMs). DDR1 is a receptor tyrosine kinase activated by matrix collagens and implicated in diseases such as cancer, fibrosis and hypoxia. Herein we describe the synthesis and inhibitory activity of compounds generated from DGMs. Three compounds were found to have sub‐micromolar inhibitory activity. The most potent of which, compound 3 (N‐(4‐chloro‐3‐((pyridin‐3‐yloxy)methyl)phenyl)‐3‐(trifluoromethyl)benzamide), had an IC₅₀ value of 92.5 nM. Furthermore, these compounds were predicted to interact with DDR1, which have a desired pharmacophore derived from a known DDR1 inhibitor. The results of synthesis and experiments indicated that our de novo design strategy is practical for hit identification and scaffold hopping.

The goal of drug design is to discover molecular structures that have suitable pharmacological properties in vast chemical space. In recent years, the use of deep generative models (DGMs) is getting a lot of attention as an effective method of generating new molecules with desired properties. However, most of the properties do not have three-dimensional (3D) information, such as shape and pharmacophore. In drug discovery, pharmacophores are valuable clues in finding active compounds. In this study, we propose a computational strategy based on deep reinforcement learning for generating molecular structures with a desired pharmacophore. In addition, to extract selective molecules against a target protein, chemical genomics-based virtual screening (CGBVS) is used as post-processing method of deep reinforcement learning. As an example study, we have employed this strategy to generate molecular structures of selective TIE2 inhibitors. This strategy can be adopted into general use for generating selective molecules with a desired pharmacophore.

Prediction of molecular properties plays a critical role towards rational drug design. In this study, the Molecular Topographic Map (MTM) is proposed, which is a two-dimensional (2D) map that can be used to represent a molecule. An MTM is generated from the atomic features set of a molecule using generative topographic mapping and is then used as input data for analyzing structure-property/activity relationships. In the visualization and classification of 20 amino acids, differences of the amino acids can be visually confirmed from and revealed by hierarchical clustering with a similarity matrix of their MTMs. The prediction of molecular properties was performed on the basis of convolutional neural networks using MTMs as input data. The performance of the predictive models using MTM was found to be equal to or better than that using Morgan fingerprint or MACCS keys. Furthermore, data augmentation of MTMs using mixup has improved the prediction performance. Since molecules converted to MTMs can be treated like 2D images, they can be easily used with existing neural networks for image recognition and related technologies. MTM can be effectively utilized to predict molecular properties of small molecules to aid drug discovery research.

Deep machine learning is expanding the conceptual framework and capacity of computational compound design, enabling new applications through generative modeling. We have explored the systematic design of covalent protein kinase inhibitors by learning from kinome-relevant chemical space, followed by focusing on an exemplary kinase of interest. Covalent inhibitors experience a renaissance in drug discovery, especially for targeting protein kinases. However, computational design of this class of inhibitors has thus far only been little investigated. To this end, we have devised a computational approach combining fragment-based design and deep generative modeling augmented by three-dimensional pharmacophore screening. This approach is thought to be particularly relevant for medicinal chemistry applications because it combines knowledge-based elements with deep learning and is chemically intuitive. As an exemplary application, we report for Bruton’s tyrosine kinase (BTK), a major drug target for the treatment of inflammatory diseases and leukemia, the generation of novel candidate inhibitors with a specific chemically reactive group for covalent modification, requiring only little target-specific compound information to guide the design efforts. Newly generated compounds include known inhibitors and characteristic substructures and many novel candidates, thus lending credence to the computational approach, which is readily applicable to other targets.

The SAR Matrix (SARM) approach was originally conceived for the systematic identification of analog series, their structural organization, and graphical structure‐activity relationship (SAR) analysis. For structurally related series, SARMs also produce virtual candidate compounds. Hence, SARM represents a unique computational approach establishing a direct link between SAR visualization and compound design. The SARM data structure is reminiscent of R‐group tables and hence easily accessible from a chemical perspective, although the underlying algorithmic basis is complex. The SARM concept has been extended in different ways to further increase its analytical and design capacity. While the efforts were largely driven from a research perspective, they have also increased the utility for practical applications. Among others, extensions include approaches for SARM‐based compound activity prediction, the generation of a large SARM database for analog searching, and the design of a deep learning architecture for advanced analog design taking chemical space information for target families into account. Herein, the SARM approach and its extensions are discussed within their scientific context.

In drug design, the prediction of new active compounds from protein sequence data has only been attempted in a few studies thus far. This prediction task is principally challenging because global protein sequence similarity has strong evolutional and structural implications, but is often only vaguely related to ligand binding. Deep language models adapted from natural language processing offer new opportunities to attempt such predictions via machine translation by directly relating amino acid sequences and chemical structures to each based on textual molecular representations. Herein, we introduce a biochemical language model with transformer architecture for the prediction of new active compounds from sequence motifs of ligand binding sites. In a proof-of-concept application on inhibitors of more than 200 human kinases, the Motif2Mol model revealed promising learning characteristics and an unprecedented ability to consistently reproduce known inhibitors of different kinases.