Maja A. Marinovi?Edward T. PetriLjubica M. Grbovi?Bojana R. Vasiljevi?...
17页
查看更多>>摘要：Abstract Human aldo‐keto reductase 1C isoforms (AKR1C1‐C4) catalyze reduction of endogenous and exogenous compounds, including therapeutic drugs, and are associated with chemotherapy resistance. AKR1C2 is involved in metastatic processes and is a target for the treatment of various cancers. Here we used molecular docking to explore the potential of a series of eleven bile acid methyl esters as AKR1C2 inhibitors. Autodock 4.2 ranked 10 of the 11 test compounds above a decoy set generated based on ursodeoxycholic acid, a known AKR1C2 inhibitor, while 5 of these 10 ranked above 94?% of decoys in Autodock Vina. Seven inactives reported in the literature not to inhibit AKR1C2 ranked below the decoy threshold: 5 of these are specific inhibitors of AKR1C3, a related isoform. Using the same parameters, Autodock Vina identified steroidal analogs of AKR1C substrates, bile acids, and AKR1C inhibitors in the top 5?% of a virtual screen of a natural product library. In experimental assays, 6 out of 11 of the tested bile acid methyl esters inhibited >50?% of AKR1C2 activity, while 2 compounds were strong AKR1C3 inhibitors. Potential off‐target interactions with the glucocorticoid receptor were measured using a yeast‐based fluorescence assay, where results suggest that the methyl ester could interfere with binding. The top ranking compound based on docking and experimental results showed dose‐dependent inhibition of AKR1C2 with an IC50 of ～3.6?μM. Molecular dynamics simulations (20?ns) were used to explore potential interactions between a bile acid methyl ester and residues in the AKR1C2 active site. Our molecular docking results identify AKR1C2 as a target for bile acid methyl esters, which combined with virtual screening results could provide new directions for researchers interested in synthesis of AKR1C inhibitors.
查看更多>>摘要：Abstract Blood‐brain‐barrier permeability (BBBP) is an important property that is used to establish the drug‐likeness of a molecule, as it establishes whether the molecule can cross the BBB when desired. It also eliminates those molecules which are not supposed to cross the barrier, as doing so would lead to toxicity. BBBP can be measured in?vivo, in?vitro or in silico. With the advent and subsequent rise of in silico methods for virtual drug screening, quite a bit of work has been done to predict this feature using statistical machine learning (ML) and deep learning (DL) based methods. In this work a mixed DL‐based model, consisting of a Multi‐layer Perceptron (MLP) and Convolutional Neural Network layers, has been paired with Mol2vec. Mol2vec is a convenient and unsupervised machine learning technique which produces high‐dimensional vector representations of molecules and its molecular substructures. These succinct vector representations are utilized as inputs to the mixed DL model that is used for BBBP predictions. Several well‐known benchmarks incorporating BBBP data have been used for supervised training and prediction by our mixed DL model which demonstrates superior results when compared to existing ML and DL techniques used for predicting BBBP.
查看更多>>摘要：Abstract c‐Myc is a major oncogene that is estimated to result in almost all human cancers and the c‐Myc downregulation has become an attractive strategy for cancer treatment. For it is hard to design compounds that can directly interact with the c‐Myc protein, the DNA G‐quadruplex (G4) was discovered in its promoter region which was referred to as a potential drug target for controlling c‐Myc expression. In this study, a combined strategy of molecular docking‐based virtual screening, molecular dynamics (MD) simulation, and molecular mechanics/generalized Born surface area (MM/GBSA) free energy calculation was conducted on the existing FDA‐Approved Drugs Library, eight compounds were selected for further experimental assay. Among them, five compounds exhibited dose‐dependently anticancer activities against RPMI‐8226 cells with IC50 values less than 18.4?μM. Further experiments showed that Trovafloxacin, Ozanimod, and Ozenoxacin decreased c‐Myc mRNA level obviously and downregulated c‐Myc expression significantly. In summary, compounds Trovafloxacin, Ozanimod, and Ozenoxacin might be regarded as new c‐Myc G4 stabilizers for the treatment of c‐Myc related cancers in the future.
查看更多>>摘要：Abstract Quantitative structure–property relationship (QSPR) modeling was investigated to predict drug and drug‐like compounds solubility in supercritical carbon dioxide. A dataset of 148 drugdrug‐like compounds, accounting for 3971 experimental data points (EDPs), was collected and used for modelling the relationship between selected molecular descriptors and solubility fraction data achieved by a nonlinear approach (Artificial neural network, ANN) based on molecular descriptors. Experimental solubility data for a given drug were published as a function of temperature and pressure. In the present study, 11 significant PaDEL descriptors (AATS3v, MATS2e, GATS4c, GATS3v, GATS4e, GATS3?s, nBondsM, AVP‐0, SHBd, MLogP, and MLFER_S), the temperature and the pressure were statistically proved to be sufficient inputs. The architecture of the optimized model was found to be {13,10,1}. Several statistical metrics, including average absolute relative deviation (AARD=3.7748?%), root mean square error (RMSE=0.5162), coefficient of correlation (r=0.9761), coefficient of determination (R2=0.9528), and robustise (Q2=0.9528) were used to validate the obtained model. The model was also subjected to an external test by using 143 EDPs. Sensitivity analysis and domain of application were examined. The overall results confirmed that the optimized ANN‐QSPR model is suitable for the correlation and prediction of this property.
查看更多>>摘要：Abstract Identifying druggable ligand‐binding sites on the surface of the macromolecular targets is an important process in structure‐based drug discovery. Deep‐learning models have been shown to successfully predict ligand‐binding sites of proteins. As a step toward predicting binding sites in RNA and RNA‐protein complexes, we employ three‐dimensional convolutional neural networks. We introduce a dataset splitting approach to minimize structure‐related bias in training data, and investigate the influence of protein‐based neural network pre‐training before fine‐tuning on RNA structures. Models that were pre‐trained on proteins considerably outperformed the models that were trained exclusively on RNA structures. Overall, 71?% of the known RNA binding sites were correctly located within 4?? of their true centres.
NSTL