VEGF is an important signaling protein involved in both vasculogenesis and angiogenesis. As an essential receptor protein tyrosine kinase propagating cellular signal transduction processes, VEGFR-2 is a central target for drug discovery against tumor-associated angiogenesis. Since the autophosphorylation of VEGFR-2 represents a key step in this signal pathway that contributes to angiogenesis, the discovery of small molecule inhibitors that block this reaction has attracted great interest for novel drugs research and development. Advances in the understanding of catalytic cleft and the conformational changes of DFG motif have resulted in the development of small molecule inhibitors known as type I and type II. High-resolution crystal structures of various inhibitors in complex with the receptor offer an insight into the relationship among binding modes, inhibition mechanisms, activity, selectivity and resistance. To control selectivity, improve activity and introduce intellectual property novelty, the strategies for the further development are discussed through structural and conformational analysis in this review.

With the rise of artificial intelligence (AI) in drug discovery, de novo molecular generation provides new ways to explore chemical space. However, because de novo molecular generation methods rely on abundant known molecules, generated molecules may have a problem of novelty. Novelty is important in highly competitive areas of medicinal chemistry, such as the discovery of kinase inhibitors. In this study, de novo molecular generation based on recurrent neural networks was applied to discover a new chemical space of kinase inhibitors. During the application, the practicality was evaluated, and new inspiration was found. With the successful discovery of one potent Pim1 inhibitor and two lead compounds that inhibit CDK4, AI-based molecular generation shows potentials in drug discovery and development.

Introduction: Quorum sensing (QS) is a cell density-dependent phenomenon in which specific pathways are activated after autoinducers (AIs) outside the microorganism reach a threshold concentration. QS creates a positive feedback loop that induces a cascade of gene expression and causes biofilm formation, virulence and sporulation. QS signals are diverse, acyl-homoserine lactone (AHL), AI peptide (AIP) and AI-2 are three major categories of QS signals. QS inhibitors (QSIs) can disrupt or prevent the formation of biofilm and reduce virulence while exerting less selective pressure on the bacteria, suggesting that QSIs are potential alternatives for antibiotics. Areas covered: This review summarized the pertinent patents on quorum sensing inhibition available from 2014 to 2018. The authors analyze these patents and provided an overview of them and their potential applications. Expert opinion: The main strategy for QS inhibition is to use the analogues of various QS signals to block downstream signal transducers. The inactivation of signal molecules or the stimulation of the immune response are also attractive strategies to inhibit QS. However, additional clinical trials are needed to assess their efficacy in mammals. In sum, QS inhibition can reduce the virulence of bacteria without affecting their growth or killing them and the reduced pressure may minimize the increasingly resistance.

Introduction: Artificial intelligence systems based on neural networks (NNs) find rules for drug discovery according to training molecules, but first, the molecules need to be represented in certain ways. Molecular descriptors and fingerprints have been used as inputs for artificial neural networks (ANNs) for a long time, while other ways for describing molecules are used only for storing and presenting molecules. With the development of deep learning, variants of ANNs are now able to use different kinds of inputs, which provide researchers with more choices for drug discovery. Areas covered: The authors provide a brief overview of the applications of NNs in drug discovery. Combined with the characteristics of different ways for describing molecules, corresponding methods based on NNs provide new choices for drug discovery, including de novo drug design, ligand-based drug design, and receptor-based drug design. Expert opinion: Various ways for describing molecules can be inputs of NN-based models, and these models achieve satisfactory results in metrics. Although most of the models have not been widely applied and tested in practice, they can be the basis for automatic drug discovery in the future.

Fibrotic diseases have become a major cause of death in the developed world. AdipoR1 agonists are potent inhibitors of fibrotic responses. Here, we focused on the in silico identification of novel AdipoR1 peptide agonists. A homology model was constructed to predict the 3D structure of AdipoR1. By docking to known active peptides, the putative active site of the model was further explored. A virtual screening study was then carried out with a set of manually designed peptides using molecular docking. Peptides with high docking scores were then evaluated for their anti-fibrotic properties. The data indicated that the novel peptide Pep70 significantly inhibited the proliferation of hepatic stellate cells (HSC) and NIH-3T3 cells (18.33% and 27.80%) and resulted in favouring cell-cycle arrest through increasing the accumulation of cells in the G0/G1 phase by 17.08% and 15.86%, thereby reducing the cell population in the G2/M phase by 11.25% and 15.95%, respectively. Additionally, Pep70 exhibited the most marked suppression on the expression of α-smooth muscle actin (α-SMA), collagen type I alpha1 (COL1A1) and TGF-β1. Therefore, the peptide Pep70 was ultimately identified as an inhibitor of fibrotic responses and as a potential AdipoR1 agonist.

Virtual screening is an important means for lead compound discovery. The scoring function is the key to selecting hit compounds. Many scoring functions are currently available; however, there are no all-purpose scoring functions because different scoring functions tend to have conflicting results. Recently, neural networks, especially convolutional neural networks, have constantly been penetrating drug design and most CNN-based virtual screening methods are superior to traditional docking methods, such as Dock and AutoDock. CNNbased virtual screening is expected to improve the previous model of overreliance on computational chemical screening. Utilizing the powerful learning ability of neural networks provides us with a new method for evaluating compounds. We review the latest progress of CNN-based virtual screening and propose prospects.