A unified GCNN model for predicting CYP450 inhibitors by using graph convolutional neural networks with attention mechanism.

Undesirable drug-drug interactions (DDIs) may lead to serious adverse side effects when more than two drugs are administered to a patient simultaneously. One of the most common DDIs is caused by unexpected inhibition of a specific human cytochrome P450 (CYP450), which plays a dominant role in the metabolism of the co-administered drugs. Therefore, a unified and reliable method for predicting the potential inhibitors of CYP450 family is extremely important in drug development. In this work, graph convolutional neural network (GCN) with attention mechanism and 1-D convolutional neural network (CNN) were used to extract the features of CYP ligands and the binding sites of CYP450 respectively, which were then combined to establish a unified GCN-CNN (GCNN) model for predicting the inhibitors of 5 dominant CYP isoforms, i.e., 1A2, 2C9, 2C19, 2D6, and 3A4. Overall, the established GCNN model showed good performances on the test samples and achieved better performances than the recently proposed iCYP-MFE model by using the same datasets. Based on the heat-map analysis of the resulting molecular graphs, the key structural determinants of the CYP inhibitors were further explored.

Development of Simple and Accurate in Silico Ligand-Based Models for Predicting ABCG2 Inhibition.

The ATP binding cassette transporter ABCG2 is a physiologically important drug transporter that has a central role in determining the ADMET (absorption, distribution, metabolism, elimination, and toxicity) profile of therapeutics, and contributes to multidrug resistance. Thus, development of predictive in silico models for the identification of ABCG2 inhibitors is of great interest in the early stage of drug discovery. In this work, by exploiting a large public dataset, a number of ligand-based classification models were developed using partial least squares-discriminant analysis (PLS-DA) with molecular interaction field- and fingerprint-based structural description methods, regarding physicochemical and fragmental properties related to ABCG2 inhibition. An in-house dataset compiled from recently experimental studies was used to rigorously validated the model performance. The key molecular properties and fragments favored to inhibitor binding were discussed in detail, which was further explored by docking simulations. A highly informative chemical property was identified as the principal determinant of ABCG2 inhibition, which was utilized to derive a simple rule that had a strong capability for differentiating inhibitors from non-inhibitors. Furthermore, the incorporation of the rule into the best PLS-DA model significantly improved the classification performance, particularly achieving a high prediction accuracy on the independent in-house set. The integrative model is simple and accurate, which could be applied to the evaluation of drug-transporter interactions in drug development. Also, the dominant molecular features derived from the models may help medicinal chemists in the molecular design of novel inhibitors to circumvent ABCG2-mediated drug resistance.

University Students' Successive Development From Entrepreneurial Intention to Behavior: The Mediating Role of Commitment and Moderating Role of Family Support.

University students having high entrepreneurial intention while not transferring into actual entrepreneurial behavior is a contradictory issue in need of in-depth research. To explore the successive development mechanism of the entrepreneurial process, this study constructed a moderated mediation model to examine whether entrepreneurial commitment from three dimensions (affective, behavioral, and continuance) mediated the relationship between entrepreneurial intention and behavior, and whether this mediating process was moderated by family support. A survey was conducted among university students from six major universities in south China using the snowball sampling approach. A total of 469 valid responses were obtained (44.6% male and 55.4% female participants). Structural equation modeling was adopted for data analysis. According to the results of the confirmatory factor analysis, it was found that entrepreneurial intention had both direct and indirect positive effects on entrepreneurial behavior, while entrepreneurial commitment worked as the mediator, and family support moderated the relationship between entrepreneurial intention and behavior. Results indicated that entrepreneurial commitment bridged the path from entrepreneurial intention to behavior, and family support created the boundary effect. This finding highlights the importance of guiding students through entrepreneurial commitment toward entrepreneurial behavior, and pays special attention to the crucial role of family support under the national strategy.

Nondestructive characterization gender of chicken eggs by odor using SPME/GC-MS coupled with chemometrics.

It's a difficult task for researchers to identify the gender of chicken eggs by nondestructive approach in the early of incubation, which not only could reduce the cost of incubation, but also could improve the welfare of chicks. Therefore, SPME/GC-MS has been applied to investigate its potential as a nondestructive tool for characterizing the differences of odor between male and female chicken eggs during early of incubation and even before hatch. The results showed that more volatiles were found in female White leghorn eggs during early of incubation and 6,10-dimethyl-5,9-undecadien-2-one, 6-methyl-5-hepten-2-one, nonanal, decanal, octanal, 2-nonen-1-ol, etc. were important for the distinction of male and female White leghorn eggs during E1-E9 of incubation. 2-ethyl-1-hexanol; octanal, nonanal, 2,2,4-trimethyl-3-carboxyisopropyl pentanoic acid isobutyl ester; 2-nonen-1-ol, cyclopropanecarboxamide, heptadecane were correlated with gender of unhatched White leghorn, Hy-line brown and Jing fen eggs, respectively. Moreover, sex-related volatiles have been strongly influenced by incubation process and egg breed, and to be related to steroid hormone biosynthesis. What's more, this study enables us to develop a new visual for ovo sexing of chicken eggs and advances our understanding of the biological significance behind volatiles emitted from chicken eggs.

De Novo Molecular Design of Caspase-6 Inhibitors by a GRU-Based Recurrent Neural Network Combined with a Transfer Learning Approach.

Due to their potential in the treatment of neurodegenerative diseases, caspase-6 inhibitors have attracted widespread attention. However, the existing caspase-6 inhibitors showed more or less inevitable deficiencies that restrict their clinical development and applications. Therefore, there is an urgent need to develop novel caspase-6 candidate inhibitors. Herein, a gated recurrent unit (GRU)-based recurrent neural network (RNN) combined with transfer learning was used to build a molecular generative model of caspase-6 inhibitors. The results showed that the GRU-based RNN model can accurately learn the SMILES grammars of about 2.4 million chemical molecules including ionic and isomeric compounds and can generate potential caspase-6 inhibitors after transfer learning of the known 433 caspase-6 inhibitors. Based on the novel molecules derived from the molecular generative model, an optimal logistic regression model and Surflex-dock were employed for predicting and ranking the inhibitory activities. According to the prediction results, three potential caspase-6 inhibitors with different scaffolds were selected as the promising candidates for further research. In general, this paper provides an efficient combinational strategy for de novo molecular design of caspase-6 inhibitors.

One-thousandth-level laser power stabilization based on optical feedback from a well-designed high-split-ratio and nonpolarized beam splitter.

Laser power stabilization plays a significant role in atomic and molecular physics, quantum precision measurement, and optical sensing and measurement. In the classical method of using a feedback control loop to stabilize the laser power, the beam splitter is the conjunction element to connect the feedback beam inside the loop and the output beam outside the loop. The stability of its split ratio will directly affect the result of power stabilization, especially in demand of high split ratios for high-efficiency output. For the compatibility of a high split ratio and high stability in a power-stabilized system, we designed and manufactured a high-split-ratio nonpolarized plate beam splitter, whose split ratio was insensitive to variations of beam intensity, polarization, and ambient temperature. Based on the optical feedback of the designed beam splitter, the light intensity was closed-loop controlled by an acousto-optic modulator; finally, the power outside the loop was stabilized as well. The output power was stabilized at 537 mW and a 6 h long-term test was performed. The relative stability of laser power outside the loop in terms of root mean square and peak to peak was 2.72×10-4 and 1.60×10-3, respectively. The relative Allan standard deviation reached 2.78×10-5 at an average time of 200 s. These results will greatly benefit many practical fields that require laser power stabilization with high split ratios and one-thousandth-level stability.

A simple pan-specific RNN model for predicting HLA-II binding peptides.

The prediction of human leukocyte antigen (HLA) class II binding peptides plays important roles in understanding the mechanism of immune recognition and developing effective epitope-based vaccines. In this work, gated recurrent unit (GRU)-based recurrent neural network (RNN) was successfully employed to establish a pan-specific prediction model of HLA-II-binding peptides by using only the HLA and peptide sequence information. In comparison with the existing pan-specific models of HLA-II-binding peptides, the GRU-based RNN model covered a broad spectrum of HLA-II molecules including 50 HLA-DR, 47 HLA-DQ, and 19 HLA-DP molecules with peptide lengths varying from 8 to 43 mers. The results demonstrated strong discriminant capabilities of the GRU-based RNN model, of which the AUC values were 0.92, 0.88, and 0.88 for the training, validation, and test sets, respectively. Also, the GRU-based model showed state-of-the-art performances in predicting the binding peptides with the length ranging from 8-32 mers, which provides an efficient method for predicting HLA-II-binding peptides of longer lengths in comparison with the available methods. Overall, taking the advantages of the RNN architecture, the established pan-specific GRU model can be used for predicting accurately the HLA-II-binding peptides in a simple and direct manner.

Conformational transitions of caspase-6 in substrate-induced activation process explored by perturbation-response scanning combined with targeted molecular dynamics.

Caspase-6 participates in a series of neurodegenerative pathways, and has aroused widespread attentions as a promising molecular target for the treatment of neurodegeneration. Caspase-6 is a homodimer with 6 central-stranded ß-sheets and 5 α-helices in each monomer. Previous crystallographic studies suggested that the 60's, 90's and 130's helices of caspase-6 undergo a distinctive conformational transition upon substrate binding. Although the caspase-6 structures in apo and active states have been determined, the conformational transition process between the two states remains poorly understood. In this work, perturbation-response scanning (PRS) combined with targeted molecular dynamics (TMD) simulations was employed to unravel the atomistic mechanism of the dynamic conformational transitions underlying the substrate-induced activation process of caspase-6. The results showed that the conformational transition of caspase-6 from apo to active states is mainly characterized by structural rearrangements of the substrate-binding site as well as the conformational changes of 60's and 130's extended helices. The H-bond interactions between L1, 130's helix and 90's helix are proved to be key determinant factors for substrate-induced conformational transition. These findings provide valuable insights into the activation mechanism of caspase-6 as well as the molecular design of caspase-6 inhibitors.

Computational Insights into Allosteric Conformational Modulation of P-Glycoprotein by Substrate and Inhibitor Binding.

The ATP-binding cassette (ABC) transporter P-glycoprotein (P-gp) is a physiologically essential membrane protein that protects many tissues against xenobiotic molecules, but limits the access of chemotherapeutics into tumor cells, thus contributing to multidrug resistance. The atomic-level mechanism of how substrates and inhibitors differentially affect the ATP hydrolysis by P-gp remains to be elucidated. In this work, atomistic molecular dynamics simulations in an explicit membrane/water environment were performed to explore the effects of substrate and inhibitor binding on the conformational dynamics of P-gp. Distinct differences in conformational changes that mainly occurred in the nucleotide-binding domains (NBDs) were observed from the substrate- and inhibitor-bound simulations. The binding of rhodamine-123 can increase the probability of the formation of an intermediate conformation, in which the NBDs were closer and better aligned, suggesting that substrate binding may prime the transporter for ATP hydrolysis. By contrast, the inhibitor QZ-Leu stabilized NBDs in a much more separated and misaligned conformation, which may result in the deficiency of ATP hydrolysis. The significant differences in conformational modulation of P-gp by substrate and inhibitor binding provided a molecular explanation of how these small molecules exert opposite effects on the ATPase activity. A further structural analysis suggested that the allosteric communication between transmembrane domains (TMDs) and NBDs was primarily mediated by two intracellular coupling helices. Our computational simulations provide not only valuable insights into the transport mechanism of P-gp substrates, but also for the molecular design of P-gp inhibitors.

Partial Least-Squares Discriminant Analysis and Ensemble-Based Flexible Docking of PD-1/PD-L1 Inhibitors: A Pilot Study.

Although mAbs targeting the programmed cell death protein 1 (PD-1)/programmed cell death ligand 1 (PD-L1) pathway have achieved remarkable therapeutic potential against multiple types of cancer, it is still of great interest for researchers to develop small-molecule PD-1/PD-L1 inhibitors without the mAb-related disadvantages of no oral bioavailability and poor solid tumor penetration. However, targeting the PD-1/PD-L1 pathway with small molecules is normally considered challenging because of the flat and large interaction surface of the PD-1/PD-L1 complex. In this paper, a total of 2558 PD-1/PD-L1 inhibitors were compiled from recent patents and literatures and then used for exploring the chemical space and structural features of PD-1/PD-L1 inhibitors by partial least-squares discriminant analysis. The results showed that intramolecular H bond, amphotericity indices, radius of gyration, nonbond electrostatic energy, fractional van der Waals surface area of H-bond donors, octanol-water partition coefficient, and molecular weight are the seven key features discriminating the PD-1/PD-L1 inhibitors from noninhibitors, with the prediction accuracy larger than 0.90. Based on the seven crystal structures of the PD-L1 dimer complexed with the patent Bristol Myers Squibb (BMS) inhibitors, the feasibility of molecular docking for this unconventional binding pocket was further investigated. The results showed that the ensemble-based flexible docking protocol can reproduce the near-native binding conformations of the BMS inhibitors with a strong correlation between the IC50 values and ligand-receptor interaction energies (R = 0.81). In general, this paper delineates, for the first time, the characteristic features of the PD-1/PD-L1 inhibitors as well as a high-quality flexible docking strategy for the unconventional binding pocket of the PD-L1 dimer.

A Pan-Specific GRU-Based Recurrent Neural Network for Predicting HLA-I-Binding Peptides.

Human leukocyte antigens (HLAs) play a critical role in human-acquired immune responses by the recognition of non-self-peptides derived from exogenous bacteria, fungi, virus, and so forth. The accurate prediction of HLA-binding peptides is thus extremely useful for the mechanistic research of cell-mediated immunity and related epitope-based vaccine design. In this work, a simple pan-specific gated recurrent unit (GRU)-based recurrent neural network model was successfully proposed for predicting HLA-I-binding peptides. In comparison with the available six allele-specific, four pan-specific, and two ensemble-based prediction models, the GRU model achieves the highest area under the receiver operating characteristic curve (AUC) scores for 21 of 64 entries of the test benchmark datasets. Besides, the GRU model also achieves satisfactory performance on other 24 entries, of which the AUC scores differ by less than 0.1 from the highest scores. Overall, taking the advantages of the GRU network and auto-embedding techniques into account, the established pan-specific GRU model is more simple and direct and shows satisfactory prediction performance for HLA-I-binding peptides with varying lengths.

In Silico Prediction of the Dissociation Rate Constants of Small Chemical Ligands by 3D-Grid-Based VolSurf Method.

Accumulated evidence suggests that binding kinetic properties-especially dissociation rate constant or drug-target residence time-are crucial factors affecting drug potency. However, quantitative prediction of kinetic properties has always been a challenging task in drug discovery. In this study, the VolSurf method was successfully applied to quantitatively predict the koff values of the small ligands of heat shock protein 90α (HSP90α), adenosine receptor (AR) and p38 mitogen-activated protein kinase (p38 MAPK). The results showed that few VolSurf descriptors can efficiently capture the key ligand surface properties related to dissociation rate; the resulting models demonstrated to be extremely simple, robust and predictive in comparison with available prediction methods. Therefore, it can be concluded that the VolSurf-based prediction method can be widely applied in the ligand-receptor binding kinetics and de novo drug design researches.

A novel microRNA, SlymiR208, promotes leaf senescence via regulating cytokinin biosynthesis in tomato.

Leaf senescence is a highly-programmed developmental process during the plant life cycle. Cytokinin (CK) has been widely acknowledged as a negative regulator to delay leaf senescence. MiRNAs play key roles in a variety of developmental and physiological processes through negatively regulating their target gene expression. However, to date, the roles of microRNAs (miRNAs) in CK biosynthesis remain unclear, and the knowledge on miRNA regulation of leaf senescence is still very limited. Isopentenyltransferases (IPTs) catalyze the initial and rate-limiting step of CK biosynthesis in higher plants. Our previous work uncovered that silencing of SlIPT4 expression in tomato resulted in premature leaf senescence. Here, we identified a novel tomato miRNA, SlymiR208, which regulates the expression of SlIPT2 and SlIPT4 at the post-transcriptional level. SlymiR208 expression is ubiquitous in tomato and exhibits an opposite transition to its target transcripts in aged leaf. SlymiR208 overexpression in tomato sharply reduced the transcript levels of SlIPT2 and SlIPT4, and the concentrations of endogenous CKs in leaves. The early leaf senescence caused by SlymiR208 overexpression was consistent with the phenotype of SlIPT4-silenced lines. The data demonstrated that SlymiR208 is a positive regulator in leaf senescence through negatively regulating CK biosynthesis via targeting SlIPT2 and SlIPT4 in tomato. This study indicated that post-transcriptional regulation via miRNA is a control point of CK biosynthesis and added a new layer to the understanding of the regulation of CK biosynthesis in tomato and a new factual proof to support that miRNAs are involved in leaf senescence.

SlCAND1, encoding cullin-associated Nedd8-dissociated protein 1, regulates plant height, flowering time, seed germination, and root architecture in tomato.

KEY MESSAGE: Silencing of SlCAND1 expression resulted in dwarfish, loss of apical dominance, early flowering, suppression of seed germination, and abnormal root architecture in tomato Cullin-RING E3 ligases (CRLs)-dependent ubiquitin proteasome system mediates degradation of numerous proteins that controls a wide range of developmental and physiological processes in eukaryotes. Cullin-associated Nedd8-dissociated protein 1 (CAND1) acts as an exchange factor allowing substrate recognition part exchange and plays a vital role in reactivating CRLs. The present study reports on the identification of SlCAND1, the only one CAND gene in tomato. SlCAND1 expression is ubiquitous and positively regulated by multiple plant hormones. Silencing of SlCAND1 expression using RNAi strategy resulted in a pleiotropic and gibberellin/auxin-associated phenotypes, including dwarf plant with reduced internode length, loss of apical dominance, early flowering, low seed germination percentage, delayed seed germination speed, short primary root, and increased lateral root proliferation and elongation. Moreover, application of exogenous GA3 or IAA could partly rescue some SlCAND1-silenced phenotypes, and the expression levels of gibberellin/auxin-related genes were altered in SlCAND1-RNAi lines. These facts revealed that SlCAND1 is required for gibberellin/auxin-associated regulatory network in tomato. Although SlCAND1 is crucial for multiple developmental processes during vegetative growth stage, SlCAND1-RNAi lines didn't exhibit visible effect on fruit development and ripening. Meanwhile, we discussed that multiple physiological functions of SlCAND1 in tomato are different to previous report of its ortholog in Arabidopsis. Our study adds a new perspective on the functional roles of CAND1 in plants, and strongly supports the hypothesis that CAND1 and its regulated ubiquitin proteasome system are pivotal for plant vegetative growth but possibly have different roles in diverse plant species.

Computational insight into the conformational transition of human toll-like receptor 8 in the agonist-induced activation processes.

Communicated by Ramaswamy H. Sarma.

Subtle differences in chemical pattern between human toll-like receptor 8 agonists and antagonists: Emerging chemical patterns analysis.

Due to the potencies in the treatments of cancer, infectious diseases, and autoimmune diseases, the developments of human TLR8 (hTLR8) agonists and antagonists have attracted widespread attentions. The hTLR8 agonists and antagonists have similar structures but with completely opposite biological effects. Up to date, the subtle differences in the structures between the hTLR8 agonists and antagonists are still unknown. In this work, emerging chemical pattern (ECP) was successfully used to extract the key chemical patterns of the hTLR8 agonists and antagonists. By using CAEP classifier, an optimal ECP model with only 3 descriptors was established with the overall prediction accuracy larger than 90%. Further hierarchical cluster analysis and molecular docking showed that the H-bond and hydrophobic properties are the key features distinguishing the hTLR8 agonists from antagonists. Comparing with the antagonists, the agonists show stronger specific H-bond properties, while antagonists have stronger non-specific hydrophobic properties. The significant differences in the structural properties may be closely related to the activation/inhibition mechanism of hTLR8.

Protein-ligand interaction fingerprints for accurate prediction of dissociation rates of p38 MAPK Type II inhibitors.

Binding/unbinding kinetics are key determinants of drug potencies. However, there are still a lot of challenges in predicting kinetic properties during early-stage drug development. In this work, position-restrained molecular dynamics simulations combined with energy decomposition were applied to extract protein-ligand interaction (PLI) fingerprints along the unbinding pathway of 20 p38 mitogen-activated protein kinase (p38 MAPK) Type II inhibitors. The results showed that the electrostatic and/or van der Waals interaction fingerprints at three key positions can be used for accurate prediction of the dissociation rate constants (koff) of p38 MAPK Type II inhibitors. The strategy proposed in this paper can provide not only an efficient method of predicting the dissociation rates of the p38 MAPK Type II inhibitors, but also the atom-level mechanism of enthalpy-driven unbinding process.

An Energetically Favorable Ligand Entrance Gate of a Multidrug Transporter Revealed by Partial Nudged Elastic Band Simulations.

P-glycoprotein (P-gp) is a multidrug transporter, which harnesses the chemical energy of ATP to power the efflux of diverse chemotherapeutics out of cells and thus contributes to the development of multidrug resistance (MDR) in cancer. It has been proved that the ligand-binding pocket of P-gp is located at the transmembrane domains (TMDs). However, the access of ligands into the binding pocket remains to be elucidated, which definitely hinder the development of P-gp inhibitors. Herein, the access pathways of a well-known substrate rhodamine-123 and a cyclopeptide inhibitor QZ-Leu were characterized by time-independent partial nudged elastic band (PNEB) simulations. The decreasing free energies along the PNEB-optimized access pathway indicated that TM4/6 cleft may be an energetically favorable entrance gate for ligand entry into the binding pocket of P-gp. The results can be reconciled with a range of experimental studies, further corroborating the reliability of the gate revealed by computational simulations. Our atomic level description of the ligand access pathway provides valuable insights into the gating mechanism for drug uptake and transport by P-gp and other multidrug transporters.

Characterization and classification of volatiles from different breeds of eggs by SPME-GC-MS and chemometrics.

Volatiles of shell eggs were identified by SPME-GC-MS to characterize and discriminate white Leghorn (W), Hy-line brown (H) and Jing fen (J) hatching eggs with comparison, principal components (PC), partial least squares (PLS), random forest classification (RFC) and canonical discriminant (CD) analyses. DVB/CAR/PDMS fiber and extraction 60 min were suited to analyze the volatiles emitted from eggs. A total of 17 or 18 volatile compounds were identified in raw shell hatching eggs, namely, nonanal, decanal and 6-methly-5-hepten-2-one were the main volatile components with contributions that over 70%. The composition and/or profile of volatile compounds from W and H eggs were much more similar than J eggs. Hexanal, decanal, 6-methly-5-hepten-2-one, heptanal, etc. have greatly contributed to the distinction of W, H and J eggs in sparse (S)-PLS and orthogonal (O)-PLS models. The accuracy of RFC and CD model were 100%, 100% (initial) and 83.3% (cross-validation), respectively. Heptanal, 6-methly-5-hepten-2-one, octanal, etc. were contributed positively to the classification of W, H, J eggs in RFC, especially for heptanal.

SMD-Based Interaction-Energy Fingerprints Can Predict Accurately the Dissociation Rate Constants of HIV-1 Protease Inhibitors.

Recent research has increasingly suggested that the crucial factors affecting drug potencies are related not only to the thermodynamic properties but also to the kinetic properties. Therefore, in silico prediction of ligand-binding kinetic properties, especially the dissociation rate constant ( koff), has aroused more and more attention. However, there are still a lot of challenges that need to be addressed. In this paper, steered molecular dynamics (SMD) combined with residue-based energy decomposition was employed to predict the dissociation rate constants of 37 HIV-1 protease inhibitors (HIV-1 PIs). For the first time, a predictive model of the dissociation rate constant was established by using the interaction-energy fingerprints sampled along the ligand dissociation pathway. On the basis of the key fingerprints extracted it can be inferred that the dissociation rates of 37 HIV-1 PIs are basically determined in the first half of the dissociation processes and that the H-bond interactions with active-site Asp25 and van der Waals interactions with flap-region Ile47 and Ile50 have important influences on the dissociation processes. In general, the strategy established in this paper can provide an efficient way for the prediction of dissociation rate constants as well as the unbinding mechanism research.