RESUMO
The rational design of chemical entities with desired properties for a specific target is a long-standing challenge in drug design. Generative neural networks have emerged as a powerful approach to sample novel molecules with specific properties, termed as inverse drug design. However, generating molecules with biological activity against certain targets and predefined drug properties still remains challenging. Here, we propose a conditional molecular generation net (CMGN), the backbone of which is a bidirectional and autoregressive transformer. CMGN applies large-scale pretraining for molecular understanding and navigates the chemical space for specified targets by fine-tuning with corresponding datasets. Additionally, fragments and properties were trained to recover molecules to learn the structure-properties relationships. Our model crisscrosses the chemical space for specific targets and properties that control fragment-growth processes. Case studies demonstrated the advantages and utility of our model in fragment-to-lead processes and multi-objective lead optimization. The results presented in this paper illustrate that CMGN has the potential to accelerate the drug discovery process.
Assuntos
Desenho de Fármacos , Descoberta de Drogas , Aprendizagem , Redes Neurais de Computação , Receptores Proteína Tirosina QuinasesRESUMO
Pegmolesatide, a synthetic, polyethylene-glycolylated, peptide-based erythropoiesis-stimulating agent (ESA), has been recently approved in China. Pegmolesatide is derived from the structure of endogenous erythropoietin (EPO), a natural product in mammals. This study compared the in vitro effects and selectivity of pegmolesatide to those of recombinant EPO and carbamylated EPO (CEPO) through computer-aided analyses and biological tests. The findings indicate that pegmolesatide exhibited the same stimulating effect on erythropoiesis as EPO with fewer side effects than EPO and CEPO.
RESUMO
The interpretation of spectral data, including mass, nuclear magnetic resonance, infrared, and ultraviolet-visible spectra, is critical for obtaining molecular structural information. The development of advanced sensing technology has multiplied the amount of available spectral data. Chemical experts must use basic principles corresponding to the spectral information generated by molecular fragments and functional groups. This is a time-consuming process that requires a solid professional knowledge base. In recent years, the rapid development of computer science and its applications in cheminformatics and the emergence of computer-aided expert systems have greatly reduced the difficulty in analyzing large quantities of data. For expert systems, however, the problem-solving strategy must be known in advance or extracted by human experts and translated into algorithms. Gratifyingly, the development of artificial intelligence (AI) methods has shown great promise for solving such problems. Traditional algorithms, including the latest neural network algorithms, have shown great potential for both extracting useful information and processing massive quantities of data. This Perspective highlights recent innovations covering all of the emerging AI-based spectral interpretation techniques. In addition, the main limitations and current obstacles are presented, and the corresponding directions for further research are proposed. Moreover, this Perspective gives the authors' personal outlook on the development and future applications of spectral interpretation.
RESUMO
Structure elucidation of unknown compounds based on nuclear magnetic resonance (NMR) remains a challenging problem in both synthetic organic and natural product chemistry. Library matching has been an efficient method to assist structure elucidation. However, it is limited by the coverage of libraries. In addition, prior knowledge such as molecular fragments is neglected. To solve the problem, we propose a conditional molecular generation net (CMGNet) to allow input of multiple sources of information. CMGNet not only uses 13C NMR spectrum data as input but molecular formulas and fragments of molecules are also employed as input conditions. Our model applies large-scale pretraining for molecular understanding and fine-tuning on two NMR spectral data sets of different granularity levels to accommodate structure elucidation tasks. CMGNet generates structures based on 13C NMR data, molecular formula, and fragment information, with a recovery rate of 94.17% in the top 10 recommendations. In addition, the generative model performed well in the generation of various classes of compounds and in the structural revision task. CMGNet has a deep understanding of molecular connectivities from 13C NMR, molecular formula, and fragments, paving the way for a new paradigm of deep learning-assisted inverse problem-solving.
RESUMO
Transition metal carbides show remarkable catalysis for MgH2, and the addition of carbon materials can attach excellent cycling stability. In this paper, Mg-doped with transition metal carbides (TiC) and graphene (G) composite (denoted as Mg-TiC-G) is designed to assess the influence of TiC and graphene on the hydrogen storage performance of MgH2. The as-prepared Mg-TiC-G samples showed favorable dehydrogenation kinetics compared to the pristine Mg system. After adding TiC and graphene, the dehydrogenation activation energy of MgH2decreases from 128.4 to 111.2 kJ mol-1. The peak desorption temperature of MgH2doped with TiC and graphene is 326.5 °C, which is 26.3 °C lower than the pure Mg. The improved dehydrogenation performance of Mg-TiC-G composites is attributed to synergistic effects between catalysis and confinement.
RESUMO
Library matching using carbon-13 nuclear magnetic resonance (13C NMR) spectra has been a popular method adopted in compound identification systems. However, the usability of existing approaches has been restricted as enlarging a library containing both a chemical structure and spectrum is a costly and time-consuming process. Therefore, we propose a fundamentally different, novel approach to match 13C NMR spectra directly against a molecular structure library. We develop a cross-modal retrieval between spectrum and structure (CReSS) system using deep contrastive learning, which allows us to search a molecular structure library using the 13C NMR spectrum of a compound. In the test of searching 41,494 13C NMR spectra against a reference structure library containing 10.4 million compounds, CReSS reached a recall@10 accuracy of 91.64% and a processing speed of 0.114 s per query spectrum. When further incorporating a filter with a molecular weight tolerance of 5 Da, CReSS achieved a new remarkable recall@10 of 98.39%. Furthermore, CReSS has potential in detecting scaffolds of novel structures and demonstrates great performance for the task of structural revision. CReSS is built and developed to bridge the gap between 13C NMR spectra and structures and could be generally applicable in compound identification.
Assuntos
Espectroscopia de Ressonância MagnéticaRESUMO
A machine learning enhanced spectrum recognition system called spectrum recognition based on computer vision (SRCV) for data extraction from previously analyzed 13C and 1H NMR spectra has been developed. The intelligent system was designed with four function modules to extract data from three areas of NMR images, including 13C and 1H chemical shifts, the integral, and the range of the shift values. During this study, three machine learning models were pretrained for number recognition, which is the key procedure for NMR data extraction. The k nearest neighbor (kNN) method was selected with optimized k (k = 4), which displayed a 100% recognition rate. Subsequently, the performance of SRCV was tested and validated to have high accuracy with a short processing time (11-21 s) for each NMR spectral image. Our spectrum recognizer enables high-throughput 13C and 1H NMR data extraction from abundant spectra in the literature and has the potential to be used for spectral database construction. In addition, the system may be applicable to be developed for data import to computer-assisted structure elucidation systems, which would automate this procedure significantly. SRCV can be accessed in GitHub (https://github.com/WJmodels/SRCV).
Assuntos
Computadores , Aprendizado de MáquinaRESUMO
An efficient one-pot synthesis of O-heterocycles or aryl ketones has been achieved using Et3SiH in the presence of InCl3 via a sequential ionic hydrogenation reaction by switching the solvent. This methodology can be used to construct C-O bonds and to prepare conjugate reduction products, including chromans, tetrahydrofurans, tetrahydropyrans, dihydroisobenzofurans, dihydrochalcones, and 1,4-diones in a facile manner. In addition, a novel plausible mechanism involving a conjugate reduction and a tandem reductive cyclization was verified by experimental investigations.
RESUMO
To reveal insights into the inhibition of BCR-ABL and its mutants, structure-based computing methods, such as docking, molecular dynamics (MD) simulation, the molecular mechanics generalized born surface area (MMGBSA), and biological characterizations, were employed to analyze two main pharmacophore zones and two related regions of imatinib derivatives. The hydrophobic and halogen interactions formed by the trifluoromethyl, as well as T-shaped π-π interactions formed by the pyrimidine, were confirmed. For the imatinib derivatives, the impacts of the amide moiety (region A) and the pyridine (region B) on the formed interactions were explored. To reveal insights into the inhibition of BCR-ABL mutants, the bioactivities of imatinib, nilotinib and flumatinib against BCR-ABL mutants were evaluated, and a point mutant (Y253F) of BCR-ABL was simulated. The results of our structure-based analysis and biological characterization of imatinib derivatives towards the inhibition of wild-type BCR-ABL and its mutants may provide new ideas for the design of imatinib analogs with potent activity.
Assuntos
Proteínas de Fusão bcr-abl/antagonistas & inibidores , Mesilato de Imatinib/uso terapêutico , Inibidores de Proteínas Quinases/uso terapêutico , Humanos , Mesilato de Imatinib/farmacologia , Inibidores de Proteínas Quinases/farmacologiaRESUMO
Developing Janus kinase 2 (JAK2) inhibitors has become a significant focus for small-molecule drug discovery programs in recent years because the inhibition of JAK2 may be an effective approach for the treatment of myeloproliferative neoplasm. Here, based on three different types of fingerprints and Extreme Gradient Boosting (XGBoost) methods, we developed three groups of models in that each group contained a classification model and a regression model to accurately acquire highly potent JAK2 kinase inhibitors from the ZINC database. The three classification models resulted in Matthews correlation coefficients of 0.97, 0.94, and 0.97. Docking methods including Glide and AutoDock Vina were employed to evaluate the virtual screening effectiveness of our classification models. The R2 of three regression models were 0.80, 0.78, and 0.80. Finally, 13 compounds were biologically evaluated, and the results showed that the IC50 values of six compounds were identified to be less than 100 nM. Among them, compound 9 showed high activity and selectivity in that its IC50 value was less than 1 nM against JAK2 while 694 nM against JAK3. The strategy developed may be generally applicable in ligand-based virtual screening campaigns.
Assuntos
Descoberta de Drogas/métodos , Janus Quinase 2/antagonistas & inibidores , Aprendizado de Máquina , Inibidores de Proteínas Quinases/farmacologia , Avaliação Pré-Clínica de Medicamentos , Interface Usuário-ComputadorRESUMO
Stimulator of interferon genes (STING) is a crucial adaptor protein that can regulate the innate immune response by inducing the secretion of type Ι interferons and other cytokines after recognizing endogenous or exogenous DNA. Due to the key role of STING in the innate immune system, the activation of STING pathway is expected to be an efficacious immunotherapeutic tactic to treat cancer. In this study, we performed a structure-activity relationship study of amidobenzimidazole monomer, led to a series of ABZI STING agonist derivatives with potent STING-activating effects. Among them, compound 72, as a representative compound, markedly activated the STING-TBK1-IRF3 signaling pathway and significantly increased the mRNA and protein levels of IFN-ß, CXCL10 and IL-6 in both WT THP-1 cells and human peripheral blood mononuclear cells (hPBMCs). In addition, it was confirmed that compound 72 was highly selective for human STING, specifically targeting human STING signaling and showing no activation of m-STING.
Assuntos
Leucócitos Mononucleares , Transdução de Sinais , Humanos , Imunidade Inata , Interferons , Relação Estrutura-AtividadeRESUMO
As glutaminase C (GAC) has become an attractive target for cancer treatment by regulating glutaminolysis, thus, interest in GAC inhibitors has risen in recent years. Herein, a potential binding subpocket comprising basic residues was identified, and through extensive structure-activity relationship studies, promising inhibitors 11 and 39 were identified with robust GAC inhibitory activity and A549 cell antiproliferative activity. X-ray crystallography of the 11-GAC and 27-GAC complexes revealed a novel binding mode against GAC. The potency of 11 and 27 against GACK320A further highlighted the importance of the binding. Notably, compounds 11 and 39 regulated the cellular metabolite, thereby increasing reactive oxygen species by blocking glutamine metabolism. Compound 11 also exhibited excellent antiproliferative activity in the A549 cell xenograft model. We further proved that 11 is a safe GAC allosteric inhibitor. A basic subpocket is proposed that might provide new strategies for the development of novel GAC inhibitors in the future.
RESUMO
Developing Bruton's tyrosine kinase (BTK) inhibitors has become a significant focus in recent years because BTK inhibition is an effective approach for the treatment of B-cell malignancies. For covalent BTK inhibitors, low oral bioavailability and low kinase selectivity remain unaddressed issues; thus, more diverse inhibitors with both novel structures and selective on target binding profiles are still needed. Here, four key regions where inhibitors bind to BTK were identified by analyzing the existing crystal structures of BTK complexes. Then, a scaffold-based molecular design work flow was established by integrating fragment-growing method, deep learning-based framework XGraphBoost and molecular docking, leading to four compounds that showed potency against BTK. Optimization of compounds 1 and 2 led to the discovery of the potent BTK inhibitor compound 42 by using in vitro potency and pharmacokinetic (PK) studies to prioritize the compounds. Compound 42 exhibited great BTK inhibition activity (IC50 = 0.7 nM) along with high oral absorption. Moreover, 42 demonstrated excellent kinase selectivity, especially over EGFR kinase, and low toxicity. In a TMD8 xenograft model, 42 significantly inhibited tumor growth (TGI = 104%) at a dosage of 50 mg/kg, indicating its potential as a novel therapeutic option for B-cell lymphomas.