Extension of multi-site analogue series with potent compounds using a bidirectional transformer-based chemical language model.

Generating potent compounds for evolving analogue series (AS) is a key challenge in medicinal chemistry. The versatility of chemical language models (CLMs) makes it possible to formulate this challenge as an off-the-beaten-path prediction task. In this work, we have devised a coding and tokenization scheme for evolving AS with multiple substitution sites (multi-site AS) and implemented a bidirectional transformer to predict new potent analogues for such series. Scientific foundations of this approach are discussed and, as a benchmark, the transformer model is compared to a recurrent neural network (RNN) for the prediction of analogues of AS with single substitution sites. Furthermore, the transformer is shown to successfully predict potent analogues with varying R-group combinations for multi-site AS having activity against many different targets. Prediction of R-group combinations for extending AS with potent compounds represents a novel approach for compound optimization.

Deep learning driven de novo drug design based on gastric proton pump structures.

Existing drugs often suffer in their effectiveness due to detrimental side effects, low binding affinity or pharmacokinetic problems. This may be overcome by the development of distinct compounds. Here, we exploit the rich structural basis of drug-bound gastric proton pump to develop compounds with strong inhibitory potency, employing a combinatorial approach utilizing deep generative models for de novo drug design with organic synthesis and cryo-EM structural analysis. Candidate compounds that satisfy pharmacophores defined in the drug-bound proton pump structures, were designed in silico utilizing our deep generative models, a workflow termed Deep Quartet. Several candidates were synthesized and screened according to their inhibition potencies in vitro, and their binding poses were in turn identified by cryo-EM. Structures reaching up to 2.10 Å resolution allowed us to evaluate and re-design compound structures, heralding the most potent compound in this study, DQ-18 (N-methyl-4-((2-(benzyloxy)-5-chlorobenzyl)oxy)benzylamine), which shows a Ki value of 47.6 nM. Further high-resolution cryo-EM analysis at 2.08 Å resolution unambiguously determined the DQ-18 binding pose. Our integrated approach offers a framework for structure-based de novo drug development based on the desired pharmacophores within the protein structure.

Motif2Mol: Prediction of New Active Compounds Based on Sequence Motifs of Ligand Binding Sites in Proteins Using a Biochemical Language Model.

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.

DeepCubist: Molecular Generator for Designing Peptidomimetics based on Complex three-dimensional scaffolds.

Mimicking bioactive conformations of peptide segments involved in the formation of protein-protein interfaces with small molecules is thought to represent a promising strategy for the design of protein-protein interaction (PPI) inhibitors. For compound design, the use of three-dimensional (3D) scaffolds rich in sp3-centers makes it possible to precisely mimic bioactive peptide conformations. Herein, we introduce DeepCubist, a molecular generator for designing peptidomimetics based on 3D scaffolds. Firstly, enumerated 3D scaffolds are superposed on a target peptide conformation to identify a preferred template structure for designing peptidomimetics. Secondly, heteroatoms and unsaturated bonds are introduced into the template via a deep generative model to produce candidate compounds. DeepCubist was applied to design peptidomimetics of exemplary peptide turn, helix, and loop structures in pharmaceutical targets engaging in PPIs.

Design of MMP-1 inhibitors via SAR transfer and experimental validation.

New matrix metalloproteinase 1 (MMP-1) inhibitors were predicted using the structure-activity relationship (SAR) transfer method based on a series of analogues of kinesin-like protein 11 (KIF11) inhibitors. Compounds 5-7 predicted to be highly potent against MMP-1 were synthesized and tested for MMP-1 inhibitory activity. Among these, compound 6 having a Cl substituent at the R1 site was found to possess ca. 3.5 times higher inhibitory activity against MMP-1 than the previously reported compound 4. The observed potency was consistent with the presence of an SAR transfer event between analogous MMP-1 and KIF11 inhibitors. Pharmacophore fitting revealed that the higher inhibitory activity of compound 6 compared to compound 4 against MMP-1 might be due to a halogen bond interaction between the Cl substituent of compound 6 and residue ARG214 of MMP-1.

Rational Strategy for Designing Peptidomimetic Small Molecules Based on Cyclic Peptides Targeting Protein-Protein Interaction between CTLA-4 and B7-1.

Currently, various pharmaceutical modalities are being developed rapidly. Targeting protein-protein interactions (PPIs) is an important objective in such development. Cyclic peptides, because they have good specificity and activity, have been attracting much attention as an alternative to antibody drugs. However, cyclic peptides involve some difficulties, such as oral availability and cell permeability. Therefore, while small-molecule drugs still present many benefits, the screening of functional small-molecule compounds targeting PPIs requires a great deal of time and effort, including structural analysis of targets and hits. In this study, we investigated a rational two-step strategy to design small-molecule compounds targeting PPIs. First, we obtained inhibitory cyclic peptides that bind to cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) by ribosomal display using PUREfrex® (PUREfrex®RD) to get structure-activity relation (SAR) information. Based on that information, we converted cyclic peptides to small molecules using PepMetics® scaffolds that can mimic the α-helix or ß-turn of the peptide. Finally, we succeeded in generating small-molecule compounds with good IC50 (single-digit µM values) against CTLA-4. This strategy is expected to be a useful approach for small-molecule design targeting PPIs, even without having structural information such as that associated with X-ray crystal structures.

Regulation of CD28 binding to SH2 domains of Grb2 and PI3K by trisubstituted carboranes for T-cell activation.

Binding of adaptor molecules, such as growth factor receptor-bound protein 2 (Grb2) and phosphoinositide 3-kinase (PI3K), to the cytoplasmic region of CD28 is critical for T-cell activation. The Src homology 2 (SH2) domains of Grb2 and PI3K interact with the cytoplasmic region, including phosphorylated Tyr, of CD28. We found that trisubstituted carboranes efficiently increased the proliferation of T cells obtained from C57BL/6 mice. The carboranes specifically increased the binding of Grb2 Src homology 2 (SH2) to CD28-derived phosphopeptide but decreased the binding of PI3K C-terminal SH2 (cSH2). Based on the crystal structures of CD28-derived phosphopeptides complexed with Grb2 SH2 and PI3K cSH2, the bound structures of compound 4 (CRL266481) were modeled to determine the molecular mechanism of the regulation.

Computational method for the systematic alignment of analogue series with structure-activity relationship transfer potential across different targets.

Lead optimization focuses on the generation of analogue series (ASs) with sustainable structure-activity relationship (SAR) progression. If roadblocks are encountered during multi-property optimization, it is often desirable to replace an AS with another containing a different core structure but having similar SAR characteristics for a given target. This process represents target-based SAR transfer. A previously unexplored question is to what extent AS-based SAR transfer events might also occur across different targets. To address this question, we have developed and applied a new computational approach to systematically search for ASs with SAR transfer potential and align qualifying series in a chemically intuitive way. The methodology relies on fragment similarity scoring in combination with dynamic programming. Our large-scale analysis has revealed that SAR transfer events across different targets are more frequently observed than one might expect, providing many opportunities for the design of new SAR transfer analogues for evolving series.

DeepAS - Chemical language model for the extension of active analogue series.

In medicinal chemistry, hit-to-lead and lead optimization efforts produce analogue series (ASs), the analysis of which is of central relevance for the exploration and exploitation of structure-activity relationships (SARs) and generation of candidate compounds. The key question in any chemical optimization effort is which analogue(s) to generate next, for which computational support is typically provided through QSAR analysis and compound potency predictions. In this study, we introduce a new chemical language model for analogue design via deep learning. For this purpose, ASs comprising active compounds are ordered according to increasing potency and the chemical language model predicts preferred R-groups for new analogues on the basis of ordered R-group sequences. Hence, consistent with the principles of deep models for natural language processing, analogues with new R-groups are predicted based upon conditional probabilities taking preceding groups into account. This implicitly accounts for the potency gradient captured by an AS and detectable SAR trends, providing a new concept for analogue design. Herein, we report the AS-based chemical language model, its initial evaluation, and exemplary applications.

Development of Low-Molecular-Weight Compounds Targeting the Cancer-Associated KLF5 Transcription Factor.

Krüppel-like factor 5 (KLF5) is a potential target for anticancer drugs. However, as an intrinsically disordered protein (IDP) whose tertiary structure cannot be solved, innovative strategies are needed. We focused on its hydrophobic α-helix structure, defined as an induced helical motif (IHM), which is a possible interface for protein-protein interaction. Using mathematical analyses predicting the α-helix's structure and hydrophobicity, a 4-amino-acid site (V-A-I-F) was identified as an IHM. Low-molecular-weight compounds that mimic the main chain conformation of the α-helix with the four side chains of V-A-I-F were synthesized using bicyclic pyrazinooxadiazine-4,7-dione. These compounds selectively suppressed the proliferation and survival of cancer cells but not noncancer cells and decreased the protein but not mRNA levels of KLF5 in addition to reducing proteins of Wnt signaling. The compounds further suppressed transplanted colorectal cancer cells in vivo without side effects. Our approach appears promising for developing drugs against key IDPs.

Approach for the Design of Covalent Protein Kinase Inhibitors via Focused Deep Generative Modeling.

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.

Comprehensive exploration of chemical space using trisubstituted carboranes.

A total of 42 trisubstituted carboranes categorised into five scaffolds were systematically designed and synthesized by exploiting the different reactivities of the twelve vertices of o-, m-, and p-carboranes to cover all directions in chemical space. Significant inhibitors of hypoxia inducible factor transcriptional activitay were mainly observed among scaffold V compounds (e.g., Vi-m, and Vo), whereas anti-rabies virus activity was observed among scaffold V (Va-h), scaffold II (IIb-g), and scaffold IV (IVb) compounds. The pharmacophore model predicted from compounds with scaffold V, which exhibited significant anti-rabies virus activity, agreed well with compounds IIb-g with scaffold II and compound IVb with scaffold IV. Normalized principal moment of inertia analysis indicated that carboranes with scaffolds I-V cover all regions in the chemical space. Furthermore, the first compounds shown to stimulate the proliferation of the rabies virus were found among scaffold V carboranes.

Visualized and Quantitative Conformational Analysis of Peptidomimetics.

Protein-protein interactions (PPIs) are fundamentally important and challenging drug targets. Peptidomimetic molecules of various types have been developed to modulate PPIs. A particularly promising drug discovery strategy, structural peptidomimetics, was designed based on special mimicking of side-chain Cα-Cß bonds. It is simple and versatile. Nevertheless, no quantitative method has been established to evaluate its similarity to a target peptide motif. We developed two methods that enable visual, comprehensive, and quantitative analysis of peptidomimetics: peptide conformation distribution (PCD) plot and peptidomimetic analysis (PMA) map. These methods specifically examine multiple side-chain Cα-Cß bonds of a peptide fragment motif and their corresponding bonds (pseudo-Cα-Cß bonds) in a mimetic molecule instead of φ and ψ angles of a single amino acid in the traditional Ramachandran plot. The PCD plot is an alignment-free method, whereas the PMA map is an alignment-based method providing distinctive and complementary analysis. Results obtained from analysis using these two methods indicate our multifacial α-helix mimetic scaffold 12 as an excellent peptidomimetic that can precisely mimic the spatial positioning of side-chain functional groups of α-helix. These methods are useful for visualized and quantified evaluation of peptidomimetics and for the rational design of new mimetic scaffolds.

Computational Prediction of Compound-Protein Interactions for Orphan Targets Using CGBVS.

A variety of Artificial Intelligence (AI)-based (Machine Learning) techniques have been developed with regard to in silico prediction of Compound-Protein interactions (CPI)-one of which is a technique we refer to as chemical genomics-based virtual screening (CGBVS). Prediction calculations done via pairwise kernel-based support vector machine (SVM) is the main feature of CGBVS which gives high prediction accuracy, with simple implementation and easy handling. We studied whether the CGBVS technique can identify ligands for targets without ligand information (orphan targets) using data from G protein-coupled receptor (GPCR) families. As the validation method, we tested whether the ligand prediction was correct for a virtual orphan GPCR in which all ligand information for one selected target was omitted from the training data. We have specifically expressed the results of this study as applicability index and developed a method to determine whether CGBVS can be used to predict GPCR ligands. Validation results showed that the prediction accuracy of each GPCR differed greatly, but models using Multiple Sequence Alignment (MSA) as the protein descriptor performed well in terms of overall prediction accuracy. We also discovered that the effect of the type compound descriptors on the prediction accuracy was less significant than that of the type of protein descriptors used. Furthermore, we found that the accuracy of the ligand prediction depends on the amount of ligand information with regard to GPCRs related to the target. Additionally, the prediction accuracy tends to be high if a large amount of ligand information for related proteins is used in the training.

Trimebutine suppresses Toll-like receptor 2/4/7/8/9 signaling pathways in macrophages.

Because of the critical roles of Toll-like receptors (TLRs) and receptor for advanced glycation end-products (RAGE) in the pathophysiology of various acute and chronic inflammatory diseases, continuous efforts have been made to discover novel therapeutic inhibitors of TLRs and RAGE to treat inflammatory disorders. A recent study by our group has demonstrated that trimebutine, a spasmolytic drug, suppresses the high mobility group box 1‒RAGE signaling that is associated with triggering proinflammatory signaling pathways in macrophages. Our present work showed that trimebutine suppresses interleukin-6 (IL-6) production in lipopolysaccharide (LPS, a stimulant of TLR4)-stimulated macrophages of RAGE-knockout mice. In addition, trimebutine suppresses the LPS-induced production of various proinflammatory cytokines and chemokines in mouse macrophage-like RAW264.7 cells. Importantly, trimebutine suppresses IL-6 production induced by TLR2-and TLR7/8/9 stimulants. Furthermore, trimebutine greatly reduces mortality in a mouse model of LPS-induced sepsis. Studies exploring the action mechanism of trimebutine revealed that it inhibits the LPS-induced activation of IL-1 receptor-associated kinase 1 (IRAK1), and the subsequent activations of extracellular signal-related kinase 1/2 (ERK1/2), c-Jun N-terminal kinase (JNK), and nuclear factor-κB (NF-κB). These findings suggest that trimebutine exerts anti-inflammatory effects on TLR signaling by downregulating IRAK1‒ERK1/2‒JNK pathway and NF-κB activity, thereby indicating the therapeutic potential of trimebutine in inflammatory diseases. Therefore, trimebutine can be a novel anti-inflammatory drug-repositioning candidate and may provide an important scaffold for designing more effective dual anti-inflammatory drugs that target TLR/RAGE signaling.

Prediction of Molecular Properties Using Molecular Topographic Map.

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.

Development of curcumin-based amyloid ß aggregation inhibitors for Alzheimer's disease using the SAR matrix approach.

Amyloid ß (Aß) aggregation inhibitor activity cliff involving a curcumin structure was predicted using the SAR Matrix method on the basis of 697 known Aß inhibitors from ChEMBL (data set 2487). Among the compounds predicted, compound B was found to possess approximately 100 times higher inhibitory activity toward Aß aggregation than curcumin. TEM images indicate that compound B induced the shortening of Aß fibrils and increased the generation of Aß oligomers in a concentration dependent manner. Furthermore, compound K, in which the methyl ester of compound B was replaced by the tert-butyl ester, possessed low cytotoxicity on N2A cells and attenuated Aß-induced cytotoxicity, indicating that compound K would have an ability for preventing neurotoxicity caused by Aß aggregation.

Discovery of Novel eEF2K Inhibitors Using HTS Fingerprint Generated from Predicted Profiling of Compound-Protein Interactions.

Background: Eukaryotic elongation factor 2 kinase (eEF2K) regulates the elongation stage of protein synthesis by phosphorylating eEF2, a process related to various diseases including cancer and cardiovascular and neurodegenerative diseases. In this study, we describe the identification of novel eEF2K inhibitors using high-throughput screening fingerprints (HTSFP) generated from predicted profiling of compound-protein interactions (CPIs). Methods: We utilized computationally generated HTSFPs referred to as chemical genomics-based fingerprint (CGBFP). Generally, HTSFPs are generated from multiple biochemical or cell-based assay data. On the other hand, CGBFPs are generated from computational prediction of CPIs using the Chemical Genomics-Based Virtual Screening (CGBVS) method. Therefore, CGBFPs do not have missing information mainly caused by the absence of assay data. Results: Chemogenomics-Based Similarity Profiling (CGBSP) of the screening library (2.6 million compounds) yielded 27 compounds which were evaluated for in vitro eEF2K inhibitory activity. Three compounds with interesting results were identified. Compounds 2 (IC50 = 11.05 µM) and 4 (IC50 = 43.54 µM) are thieno[2,3-b]pyridine derivatives that have the same scaffolds with a known eEF2K inhibitor, while compound 13 (IC50 = 70.13 µM) was a new thiophene-2-amine-type eEF2K inhibitor. Conclusions: CGBSP supplied an efficient strategy in the identification of novel eEF2K inhibitors and provided useful scaffolds for optimization.