Mingyue Zheng

Bio: Mingyue Zheng is an academic researcher from Chinese Academy of Sciences. The author has contributed to research in topics: Medicine & Virtual screening. The author has an hindex of 34, co-authored 215 publications receiving 3867 citations. Previous affiliations of Mingyue Zheng include Ocean University of China & Southern Medical University.

Pushing the Boundaries of Molecular Representation for Drug Discovery with the Graph Attention Mechanism.

Abstract: Hunting for chemicals with favorable pharmacological, toxicological, and pharmacokinetic properties remains a formidable challenge for drug discovery. Deep learning provides us with powerful tools to build predictive models that are appropriate for the rising amounts of data, but the gap between what these neural networks learn and what human beings can comprehend is growing. Moreover, this gap may induce distrust and restrict deep learning applications in practice. Here, we introduce a new graph neural network architecture called Attentive FP for molecular representation that uses a graph attention mechanism to learn from relevant drug discovery data sets. We demonstrate that Attentive FP achieves state-of-the-art predictive performances on a variety of data sets and that what it learns is interpretable. The feature visualization for Attentive FP suggests that it automatically learns nonlocal intramolecular interactions from specified tasks, which can help us gain chemical insights directly from data beyond human perception.

In silico ADME/T modelling for rational drug design

Abstract: In recent decades, in silico absorption, distribution, metabolism, excretion (ADME), and toxicity (T) modelling as a tool for rational drug design has received considerable attention from pharmaceutical scientists, and various ADME/T-related prediction models have been reported. The high-throughput and low-cost nature of these models permits a more streamlined drug development process in which the identification of hits or their structural optimization can be guided based on a parallel investigation of bioavailability and safety, along with activity. However, the effectiveness of these tools is highly dependent on their capacity to cope with needs at different stages, e.g. their use in candidate selection has been limited due to their lack of the required predictability. For some events or endpoints involving more complex mechanisms, the current in silico approaches still need further improvement. In this review, we will briefly introduce the development of in silico models for some physicochemical parameters, ADME properties and toxicity evaluation, with an emphasis on the modelling approaches thereof, their application in drug discovery, and the potential merits or deficiencies of these models. Finally, the outlook for future ADME/T modelling based on big data analysis and systems sciences will be discussed.

Preclinical characterization of anlotinib, a highly potent and selective vascular endothelial growth factor receptor-2 inhibitor.

Abstract: Abrogating tumor angiogenesis by inhibiting vascular endothelial growth factor receptor-2 (VEGFR2) has been established as a therapeutic strategy for treating cancer. However, because of their low selectivity, most small molecule inhibitors of VEGFR2 tyrosine kinase show unexpected adverse effects and limited anticancer efficacy. In the present study, we detailed the pharmacological properties of anlotinib, a highly potent and selective VEGFR2 inhibitor, in preclinical models. Anlotinib occupied the ATP-binding pocket of VEGFR2 tyrosine kinase and showed high selectivity and inhibitory potency (IC50 <1 nmol/L) for VEGFR2 relative to other tyrosine kinases. Concordant with this activity, anlotinib inhibited VEGF-induced signaling and cell proliferation in HUVEC with picomolar IC50 values. However, micromolar concentrations of anlotinib were required to inhibit tumor cell proliferation directly in vitro. Anlotinib significantly inhibited HUVEC migration and tube formation; it also inhibited microvessel growth from explants of rat aorta in vitro and decreased vascular density in tumor tissue in vivo. Compared with the well-known tyrosine kinase inhibitor sunitinib, once-daily oral dose of anlotinib showed broader and stronger in vivo antitumor efficacy and, in some models, caused tumor regression in nude mice. Collectively, these results indicate that anlotinib is a well-tolerated, orally active VEGFR2 inhibitor that targets angiogenesis in tumor growth, and support ongoing clinical evaluation of anlotinib for a variety of malignancies.

TransformerCPI: improving compound-protein interaction prediction by sequence-based deep learning with self-attention mechanism and label reversal experiments.

Abstract: Motivation Identifying compound-protein interaction (CPI) is a crucial task in drug discovery and chemogenomics studies, and proteins without three-dimensional structure account for a large part of potential biological targets, which requires developing methods using only protein sequence information to predict CPI. However, sequence-based CPI models may face some specific pitfalls, including using inappropriate datasets, hidden ligand bias and splitting datasets inappropriately, resulting in overestimation of their prediction performance. Results To address these issues, we here constructed new datasets specific for CPI prediction, proposed a novel transformer neural network named TransformerCPI, and introduced a more rigorous label reversal experiment to test whether a model learns true interaction features. TransformerCPI achieved much improved performance on the new experiments, and it can be deconvolved to highlight important interacting regions of protein sequences and compound atoms, which may contribute chemical biology studies with useful guidance for further ligand structural optimization. Availability and implementation https://github.com/lifanchen-simm/transformerCPI.

Pattern Recognition and Machine Learning

Abstract: Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Cu-catalyzed azide-alkyne cycloaddition.

Abstract: The Huisgen 1,3-dipolar cycloaddition reaction of organic azides and alkynes has gained considerable attention in recent years due to the introduction in 2001 of Cu(1) catalysis by Tornoe and Meldal, leading to a major improvement in both rate and regioselectivity of the reaction, as realized independently by the Meldal and the Sharpless laboratories. The great success of the Cu(1) catalyzed reaction is rooted in the fact that it is a virtually quantitative, very robust, insensitive, general, and orthogonal ligation reaction, suitable for even biomolecular ligation and in vivo tagging or as a polymerization reaction for synthesis of long linear polymers. The triazole formed is essentially chemically inert to reactive conditions, e.g. oxidation, reduction, and hydrolysis, and has an intermediate polarity with a dipolar moment of ∼5 D. The basis for the unique properties and rate enhancement for triazole formation under Cu(1) catalysis should be found in the high ∆G of the reaction in combination with the low character of polarity of the dipole of the noncatalyzed thermal reaction, which leads to a considerable activation barrier. In order to understand the reaction in detail, it therefore seems important to spend a moment to consider the structural and mechanistic aspects of the catalysis. The reaction is quite insensitive to reaction conditions as long as Cu(1) is present and may be performed in an aqueous or organic environment both in solution and on solid support.